Citation
Making Revit User Friendly and More Intuitive by Adding Guidance And Query Chatbot

Material Information

Title:
Making Revit User Friendly and More Intuitive by Adding Guidance And Query Chatbot
Creator:
Waked, Maroun
Publisher:
University of Florida
Publication Date:
Language:
English

Thesis/Dissertation Information

Degree:
Master's ( M.S.C.M)
Degree Grantor:
University of Florida
Degree Disciplines:
Construction Management
Committee Chair:
Issa,Raja Raymond
Committee Co-Chair:
Muszynski,Larry C
Committee Members:
Alwisy,Aladdin
Graduation Date:
12/18/2020

Subjects

Subjects / Keywords:
add-in
autodesk
bim
bot
chatbot
ibm-watson
plugin
revit
revit-api

Notes

General Note:
Software for the construction industry has progressed in sophistication over the last three decades, with advances seen in cost and ease-of-use. Moreover, most software programs allow users the ability to create custom tools either using scripting or accessing the application programming interface. AutodeskTM Revit is a common building information modeling (BIM) program used by many people in the architecture/engineering/construction industry. In this review, two types of chatbots were created in order to make AutodeskTM Revit a more intuitive and user-friendly software. Different types of users would benefit from a AutodeskTM Revit add-in in the form of a chatbot that would make them more accurate, fast, knowledgeable, and efficient for various types of tasks. One scenario is that they must create a bill of quantity (BOQ). During the process, the user needs to create either a Schedule of Quantity or a Schedule of Material for each item needed which is a tedious repetitive work. In this case, the Query chatbot which is a program that could complete a quantity takeoff process automatically and could increase productivity and ease of work. Another scenario could be when the user is not sure of what tool to is required. In this case, Guiding chatbot is developed which is a program that could guide the user to what tool to use and where it is located. Creating a AutodeskTM Revit tool that can inform the user on how to do a certain task would help advance the level (Beginner-Professional) of users.

Record Information

Source Institution:
UFRGP
Rights Management:
All applicable rights reserved by the source institution and holding location.
Embargo Date:
12/31/2022

Downloads

This item has the following downloads:


Full Text

PAGE 1

MAKING REVIT USER FRIENDLY AND MORE INTUITIVE BY ADDING GUIDANCE AND QUERY CHATBOT By MAROUN WAKED A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE IN CONSTRUCTION MANAGEMENT UNIVERSITY OF FLORIDA 2020

PAGE 2

© 2020 Maroun Waked

PAGE 3

To Labib, Rita, Rita junior and Sasha, for their endless love and support

PAGE 4

4 ACKNOWLEDGMENTS This fruitful journey would not have been possible without the support of my family, professors and mentors, and friends. I would like to express immense and special appreciation and thankfulness to my advisor, Professor Dr. R. Raymond Issa, emendous mentor for me with endless encouragement throughout my research and his advice on both my research as well as on my career have been invaluable. To my family, thank you for encouraging me in all my pursuits and inspiring me to follow my dreams. I am especially grateful to my parents, who supported me emotionally and financially. I always knew that you believed in me and wanted the best for me. Thank you for teaching me that my job in life was to learn, to be happy, and to know and understand myself ; only then could I know and understand others. Last but not least, to my girlfriend, thank you for always believing in me till the end and for always motivating me, I done it without you.

PAGE 5

5 TABLE OF CONTENTS Page ACKNOWLEDGMENTS ................................ ................................ ................................ .. 4 LIST OF FIGURES ................................ ................................ ................................ .......... 7 ABSTRACT ................................ ................................ ................................ ..................... 8 CHAPTER 1 INTRODUCTION ................................ ................................ ................................ .... 10 1.1 Scope of Research ................................ ................................ ........................... 10 1.1.1 Aim ................................ ................................ ................................ .......... 10 1.1.2 Research Objectives ................................ ................................ ............... 11 1.1.3 Limitations ................................ ................................ ............................... 11 1.2 Statement of Purpose ................................ ................................ ....................... 12 2 LITERATURE REVIEW ................................ ................................ .......................... 14 2.1 AutodeskTM Revit ................................ ................................ ............................. 14 2.2 Chatbot ................................ ................................ ................................ ............. 15 2.3 IBM Watson ................................ ................................ ................................ ...... 16 2.4 Building Information Modeling Plugins ................................ .............................. 17 2.4.1 Environmental Plugins ................................ ................................ ............. 17 2.4.2 Facility Management BIM ................................ ................................ ........ 19 2.4.3 Virtual Reality (VR) ................................ ................................ .................. 21 2.4.4 Tunneling Design ................................ ................................ .................... 22 2.4.5 Automatic Generation of Path Networks for Evacuation .......................... 23 2.5 Chatbots in Different Industries ................................ ................................ ......... 24 2.5.1 Education ................................ ................................ ................................ 25 2.5.2 Medical ................................ ................................ ................................ .... 26 2.5.3 Tourism ................................ ................................ ................................ ... 27 2.5.4 Journalism ................................ ................................ ............................... 28 3 RESEARCH METHODOLOGY ................................ ................................ ............... 30 3.1 Chatbot #1 (Query Chatbot) ................................ ................................ .............. 31 3.1.1 Creating Revit Plugin ................................ ................................ ............... 32 3.1.2 Query Chatbot Experiment ................................ ................................ ...... 39 3.2 Chatbot #2 (Guiding Chatbot) ................................ ................................ ........... 40 3.2.1 Steps to Develop a Chatbot ................................ ................................ ..... 40 3.2.2 Conversation Concepts ................................ ................................ ........... 41 3.2.2.1 Intents and entities ................................ ................................ ......... 41 3.2.2.2 Dialog ................................ ................................ ............................. 43 3.2.2.3 Dialog node ................................ ................................ .................... 44

PAGE 6

6 3.2.2.4 Context as in a real life conversation. ................................ ............ 45 3.2.2.5 Condition and responses ................................ ............................... 46 3.2.2.6 Conversation turn ................................ ................................ ........... 4 6 3.2.2.7 Typical conversation flow ................................ ............................... 47 3.2.3 Guiding Chatbot Experiment ................................ ................................ ... 47 3.3 Conclusion ................................ ................................ ................................ ........ 48 4 RESULTS AND ANALYSIS ................................ ................................ .................... 49 4.1 Experimental Metrics ................................ ................................ ........................ 49 4.2 Data Collection ................................ ................................ ................................ . 50 4.2.1 Data Collection #1: Guiding Chatbot ................................ ....................... 50 4.2.2 Data Collection #2: Query Chatbot ................................ .......................... 50 4.3 Results and Analysis ................................ ................................ ......................... 50 4.3.1 Results and Analysis #1: Guiding Chatbot ................................ .............. 50 4.3.2 Results and Analysis #2: Query Chatbot ................................ ................. 53 5 CONCLUSION AND RECOMMENDATIONS ................................ ......................... 57 5.1 Conclusion ................................ ................................ ................................ ........ 57 5.2 Recommendations ................................ ................................ ............................ 58 APPENDIX CODES SCRIPT ................................ ................................ ..................... 59 LIST OF REFERENCES ................................ ................................ ............................... 64 BIOGRAPHICAL SKETCH ................................ ................................ ............................ 72

PAGE 7

7 LIST OF FIGURES Figure page 3 1 Rational Logic and Workflow ................................ ................................ .............. 31 3 2 Project Set up and Configuration ................................ ................................ ........ 33 3 3 Adding References to the Project ................................ ................................ ....... 34 3 4 Adding Namespaces to access required contents ................................ .............. 34 3 5 Code Script Example ................................ ................................ .......................... 36 3 6 Manifest File Script Example ................................ ................................ .............. 37 3 7 Activating the Start External ................................ ................................ ............... 38 3 8 Testing the Program ................................ ................................ ........................... 39 3 9 Intents and Entities (adapted from Azraq et al. 2017) ................................ ......... 42 3 10 IBM Watson Intent Example ................................ ................................ ............... 42 3 11 IBM Watson Entity Example ................................ ................................ ............... 43 3 12 Dialog Format ................................ ................................ ................................ ..... 44 3 13 Dialog Concept (adapted from Azraq et al. 2017) ................................ ............... 45 3 14 Conversation Turn (adapted from Azraq et al. 2017) ................................ .......... 46 3 15 Conversation Flow (adapted from Azraq et al. 2017 ) ................................ ......... 47 4 1 Guiding Chatbot Speed Results ................................ ................................ ......... 51 4 2 Guiding Chatbot Experimental Results ................................ ............................... 52 4 3 Query Chatbot Speed Results ................................ ................................ ............ 54 4 4 Query Chatbot Experimental Results ................................ ................................ . 55

PAGE 8

8 Abstract of Thesis Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Science in Construction M a nagement MAKING REVIT USER FRIENDLY AND MORE INTUITIVE BY ADDING GUIDANCE AND QUERY CHATBOT By Maroun Waked December 2020 Chair: R. Raymond Issa Cochair: Larry Muszynski Major: Construction Management Software for the construction industry has progressed in sophistication over the last three decades, with advances seen in cost and ease of use. Moreover , most software programs allow users the ability to create custom tools either using scripting or acces sing the application programming interface. Autodesk TM Revit is a common building information modeling (BIM) program used by many people in the architecture/engineering/construction industry. In this review, two types of chatbots were created in order to make Autodesk TM Revit a more intuitive and user friendly software. Different types of users would benefit from a Autodesk TM Revit add in in the form of a chatbot that would make them more accurate, fast, knowledge able, and efficient for various types of tasks. One scenario is that they must create a bill of quantity (BOQ). During the process, the user needs to create either a Schedule of Quantity or a Schedule of Material for each item needed which is a tedious rep etitive work. In this case, the Query chatbot which is a program that could complete a quantity takeoff process automatically and could increase productivity and ease of work. Another scenario could be when the user is not sure of what tool to is required. In this case,

PAGE 9

9 Guiding chatbot is developed which is a program that could guide the user to what tool to use and where it is located. Creating a Autodesk TM Revit tool that can inform the user on how to do a certain task would help advance the level (Beginn er Professional) of users.

PAGE 10

10 CHAPTER 1 INTRODUCTION The world is undergoing significant transformations fueled by digitization, data and communications technology, machine learning, robotics and artificial intelligence ( Gupta et al. 2017 ). An intelligent virt ual assistant or chatbot is a software agent which utilizes Artificial Intelligence (AI), to perform tasks or services for a user depending on his/her inputs and needs. As time goes by, Organizations are using chatbots even more to offer their users a bett er customer experience and to manage chats on subjects relating to the company or its products/services. Some of the common tasks performed guiding users on how to achieve a certain task etc. (Reshmi and Balakrishnan, 2018). In the architecture, engineering, and construction (AEC) industry, building information modeling (BIM) has proved to be one of the most promising developments. With BIM technology, a precise vir tual model of a structure is digitally constructed. The primary advantage of BIM over other modeling methods is the availability and accessibility of data about the structure, which can be utilized at various stages: architectural design, building estimati ons, construction, maintenance and reconstruction. 1.1 Scope of Research 1. 1 .1 Aim Different aspects of a design and construction project can be described and illustrated using Building Information Modeling. Fundamentally, all structural, architectural and MEP components are included and coordinated into the BIM process that involves the development of an advanced 3D model for the proposed project. Being able to view those components in 3D permits project managers to have better

PAGE 11

11 coordination in the design stages through the creation of visual models of the interaction of systems with one another. This review will be concerned with the potential of making AutodeskTM Revit (BIM) a more user friendly software and more intuitive to users by implementing Chatbots into AutodeskTM Revit. The first chatbot developed is the Query Chatbot that will perform quantity takeoff. The AutodeskTM Revit API allows the Chatbot gain access to model graphical data and to model parameter data in real time. The second developed chatbot is the Guiding chatbot that guides the users on what tool to use and where it i s located. The IBM Watson platform uses Artificial Intelligence to interact with humans leading to a better user experience. The chatbots developed can assist different types of users including architects, contractors, estimators, field engineers, and elec trical engineers with different levels of proficiency. Moreover, the bots can help a fully proficient AutodeskTM Revit user specially with repetitive tasks such as BOQ and lead him to a faster end product and help a beginner level AutodeskTM Revit user un derstand any command they need to use. 1. 1 .2 Research O bjectives Apply C# programming language and the AutodeskTM Revit API to develop AutodeskTM Revit plugins . Explore the feasibility of creating a Guidance chatbot for AutodeskTM Revit that will inform an d guide the user on which tool to use and where is it located. Explore the feasibility of creating a Query chatbot as a AutodeskTM Revit add in that will help the user find, calculate or identify a certain element or material. Evaluate the performance and efficiency of the chatbot . 1. 1 .3 Limitations The limitations of this study are:

PAGE 12

12 The chatbot will be limited to the construction of buildings types facilities and wont function for infrastructure facilities. The chatbot will not be able to answer all quest ion due to the fact that each question will have its unique sample code; coming up with a sample code for each question will need an organization to work on. 1. 2 Statement of Purpose Over the last decade, BIM has evolved drastically and has become much mor e vital and necessary to be integrated into all construction phases. However, with the development of BIM software such as AutodeskTM Revit, it became so powerful and complex at the same time. Even though AutodeskTM Revit is a well known tool for its diver se aspects and performance, many construction companies are not implementing it be seen as barriers. Most of these concerns relate to the knowledge, expertise and process necessary for the implementation of BIM (Jan and Damian, 2008). Barriers for BIM adoption are at different levels and some of them are easier to remove (Enshassi et al., 2016). Some obstacles, which have influenced the slow implementation of BIM within th e construction industry are (Lee et al., 2013): Lack of knowledge regarding AutodeskTM Revit (BIM) adoption. Lack of knowledge and practices regarding the use of AutodeskTM Revit (BIM). Lack of resources both software and hardware regarding use of BIM tools. After plenty of research, Johnson and Laepple (2003) concluded that the main obstacle for BIM implementation is unwillingness of industry to shift from current practices and learnin g new technologies. The development of the chatbots will help AEC firms with implementing BIM in their workflow. By making AutodeskTM Revit more intuitive companies would fear less changing from traditional practices such as AutoCAD to new practices such as

PAGE 13

13 AutodeskTM Revit. If a design firm has a number of professional architects who have mastered the use of AutoCAD, why would the company shift to a new practice that requires time to get used to and is much more complex. In this case, the chatbots can hel p users to self learn, guide them throughout the process even if they start with no knowledge or experience about the AutodeskTM Revit software and help them with quantity takeoff. The second significant contribution of this development is that the chatbot s created would help the user to be faster and more accurate leading to a more efficient user and therefore a better productivity.

PAGE 14

14 CHAPTER 2 LITERATURE REVIEW 2.1 AutodeskTM Revit Autodesk Revit is a Building Information Modeling (BIM) software for, which allows the user to design using parametric modeling and drafting elements. Intelligent, 3D and parametric object based design are provided by Building Information Modeling (BIM) wh ich is a new Computer Aided Design (CAD) paradigm. A change anyplace is a change all over the place, instantly, with no user interaction to manually update any model. This is made conceivable by AutodeskTM Revit's underlying relational database design which is called the parametric change engine. Usually, if a part of the design is to be seen from more than one side, it will be generated in 3D and users can extract thei r own 2D and 3D entities for drawings and demonstrations. AutodeskTM Revit is a single building database that can be shared among numerous parties. Plans, sections, elevations, legends, and schedules are altogether interconnected, and if any party rolls ou t an improvement or change in any view, other perspectives and views are automatically refreshed and updated. Therefore, building objects and elements are always fully coordinated in drawings and schedules. Miniature versions of building sections can be ge nerated by either using 2D and 3D drafting objects or by importing already made drafts from other CAD platforms through SKP, SAT, DWG, DXF, or DGN. Project databases can be created from users working together on the same project and databases for verification. An interference check is performed by AutodeskTM Revit to identify whether different components of the building are occupying the same physical space (Whitleygroup, n.d.) .

PAGE 15

15 2.2 Chatbot Intelligent and self governing systems, suc h as Chatbots, are the result of new technological advances. A Chatbot is basically a computer program that essentially simulates human conversations, thus allowing a method of communication between a human and a machine. This interaction occurs in the for m of a message or voice command. A Chatbot is modified to work autonomously from a human operator (Elupula a conversation and as a software that permits an easy way of interaction between people and machines. Such interactions happen in several forms: an oral or written dialogue in a natural language, communication with devices, motion sensors, and other ways. According to Hill et al. (2015), humans can effectively adju st their language to human chatbot communication, although there are remarkable variances in the content and quality of these chats. The authors concluded that humans tend to communicate with chatbots with shorter messages, but longer periods of time as co mpared to a normal human conversation. Chatbots can be classified in two Groups: rules based and Al based. Rules based Chatbots function by means of particular commands (or keywords), which usually follow well defined navigation flows and yield targeted co nversations. Al based Chatbots utilize more advanced technologies such as machine learning, NLP, including other artifices to increase their capability for dialogue and interaction. The most important feature of a chatbot is the engine because it is the co re part responsible for transforming the common language into actions a machine can understand. Chatbots engines are typically established using numerous Natural Language

PAGE 16

16 Processing and Machine Learning models to provide satisfactory levels of accuracy (Ka r and Haldar 2016). 2.3 IBM Watson In this case, the IBM Watson Assistant engine is used for developing the Query chatbot. The IBM Watson Chatbot, a software program, uses Artificial Intelligence to interact with humans. In other words, human conversations will be stimulated wherein at one end is a machine while the other is a user. The Conversation has a feature that uses ML and NLP resources capable of obtaining intentions and entities from a chat leading to a better and specific answer. The AI based chatbot was selected due to the similar ity of experience in dialogue with a human being that it can convey. This is so, due to the NLP and ML algorithms, that allows it to draw the same significance from a dialogue, even when communicated in different ways. Technical knowledge (Resources) and e xtra knowledge (Others) are two different groups the chatbot knowledge areas were categorized into. The technical knowledge area is accountable for searching information from the WSN in the database across API, whereas the extra knowledge area was made to make chatbot performance more human like. This part is comprised of conversations like greetings, options menu and help sections, rendering the chatbot more social. A Watson Conversation Service Instance can be summoned by one or several Applications via a n API. The Watson Conversation Service also includes a modelling tool that permits iterative dialog development. As for the Cognitive feature, it is that the tool does not require to be taught with all possible alternatives and options. The invoking applic ations can also be Messaging Platforms such as Facebook Messenger and Slack. One or more Workspaces can be generated within a Service Instance and those Workspaces permit model entities, intents, and dialogs. Those

PAGE 17

17 concepts help building conversations quic kly for a Chatbot. Answering a question or processing a bill payment are Intents, reasons or objectives conveyed in a customer's input. By identifying the intent expressed in a customer's input, the Conversation service can decide the correct dialog flow f or responding to it. Entities signify a class of object or a data type that is applicable to a user's purpose. The dialog uses the intents and entities that are recognized in the user's input, in addition to the framework from the application, to interact with the user and eventually deliver a beneficial answer. The Cognitive aspect is in the fact that the tool does not need to be trained with all possible options and variations. 2.4 Building Information Modeling Plugins Architects, Engineers, Contractors a nd Designers are using and adopting AutodeskTM Revit to create a unified model consisting of real life information. AutodeskTM Revit is helping the reality based parametric modelling process, by directly manual family creation (Garagnani and Manferdini 201 3) or commercial plugins (Klein et al. 2015). Different types of plugins were developed throughout the years to expand and diversify the usability of AutodeskTM Revit. Plugins were developed to help with different aspects of the construction industry inclu ding: Environmental, Structural, Facility Management and Safety. 2.4.1 Environmental Plugins Recently, the object oriented modelling and coordination approach of BIM has faced major take up throughout the construction industry, where it utilizes virtual mo dels to create design documentation, takes off building amounts, manages construction detailing and sequencing, and connects construction data to Operation and Maintenance. In 2008, Kryiel and Nies published the first all inclusive summary about

PAGE 18

18 has focused on helping architects use BIM to make sure their designs match environmental regulations according to local codes, such as the certification of Council (Barnes and Castro code (Gandhi and Jupp 2014). The idea of Green BIM was also studied in the scope of economic as well as ecologic f actors impacting on building lifecycle analysis (Jalaei and Jrade 2014) (Akbarnezhad et al. 2014) (Wong and Zhou 2015). Additionally, not only has using sustainability tools in conjunction with BIM been explored with respect to interoperability between dif ferent BIM and analysis software applications (Kumar 2008), but also their potential of offering decision support to engineers (Inyim and Rivera 2015). Azhar and Brown (2009) give a comparison between the integration of BIM creating tools AutodeskTM Revit and ArchiCA with the already mentioned Ecotect, IE along with Green Building Studio (GBS). Nevertheless, Azhar and Brown could not present definitive results due to the ill advancement of the progression of integration of environmental analysis with BI M functions in 2009. As of 2010, Autodesk gradually incorporated parts of Ecotect into their BIM compatible authoring/massing tools like AutodeskTM Revit, Vasar or FormIT in order to increase their modelling competence with basic environmental feedback. Revi currently has many environmental testing features built into its core software, such as: Solar Analysis, Sun and Shadow Studies, Daylighting and Lighting, Thermal performance, Whole building energy analysis and Weather data visualization. In 2015, a s a direct result, AutodeskTM Ecotect was terminated as a separate product. Interest in parametric design grew among designers

PAGE 19

19 especially for form finding processes associated to early design exploration, as a parallel to the increased abundance of BIM. Mc benefited even more users who were able to establish rule models to interact with environmental analysis functions. Due to the rule based nature of parametric geometry definitions, the fast turnou t of design variations produced quick turnouts in (numerically) precise setting. Users did not need to re draw their models for every new geometric setting and parameters stay mapped across geometry and simulation when connecting parametrically defined geo metry to environmental analysis. A big advantage of the method McNeel familiarized Grashopper with Rhinocero users is that it is an open source setup and its expandability lead to the availability of a great number of low cost plugins. Just like the link s between BIM and environmental analysis, the Grasshopper community used every chance to tie parametric design to optimization. In 2009, a plugin (Geco) linking Grasshopper to AutodeskTM Ecotect was unveiled and other environmental plugins trailed such as Diva (originally developed by researchers at Harvard University), Ladybug for sun path analysis, wind roses (Roudsari and Pak 2013), or shadow studies, or (the related) Honeybee which links Grasshopper to (daylight) simulation engines such as EnergyPlu s, Radiance, or Daysim (Hu et al. 2020). 2.4.2 Facility Management BIM The use of BIM for facilities management to enable access to accurate, consistent and up to date building facility information has gained increased popularity over the last several year s. BIM could aid in facilities management in a variety of ways. Some potential application areas of the BIM and facility management integration include locating building components, enabling real time access of data, visualization and

PAGE 20

20 marketing, maintainab ility, creating digital assets, and monitoring energy (Krukowski and Arsenijevic 2010). As building facilities today are comparatively sophisticated, the requirement for facility information to maintain and operate them is of an increasing importance (Jord ani 2010). Such information can enable tracking components accurately, identifying building operations inefficiencies and simplifying the process of responding to client requests (Forns Samso 2010). Each asset associated with a facility management inventor y has a cost tied up to its installation, operations and maintenance. An accurate and updated inventory of building equipment inventory is thus essential for the budgeting of replacement and maintenance costs (GSA 2011). This process usually takes place ju st before a building is occupied, which has been identified as the best time period for collecting and uploading facility asset information. The majority of construction contracts demand a handover of those data in physical document form: including data sh eets, list of spare parts, equipment information, and warranties. This type of information, which uses design and constructional data to support operations and maintenance, is often classified as operational documents by industry professionals. Constructio n Operations Building information exchange (COBie), has gained increased popularity over the last several years as a data exchange format for facility management data handover. Within the BIM software and as a plugin COBie, sheets are usually generated in Excel format. COBie is explained as an IFC centered data exchange application which comprises data exchange between the operations and construction project stages, according to the GSA BIM facility management guide (2011). First, COBie obtains building ass et information from an AutodeskTM Revit model, then exports the maintenance plans and guidelines to an excel spreadsheet and

PAGE 21

21 finally imports this facility data into a facility engineers computerized maintenance management system. Gallaher (2004) estimated a savings of around $613 million which was identified to be spent otherwise on converting facility information into a format that is usable by facility managers. In a research project, Lavy and Jawadekar(2014), also showed that the use of BIM for data form ulation and COBie as a tool for data formatting favorably impacted facility management of certain selected projects with regard to gathering and accumulating inventory data for the purpose of preventive maintenance. Storing data extracted from BIM models a nd extracting the spatial information for integration with inventory data are important aspects of COBie. The utilization of COBie spread sheets extracted from a BIM model opened up the possibility of combining the building facility information with the ma intenance history of the components that are being repaired or maintained. COBie files are also capable of being imported and recognized by most popular CMMS software. This further facilitates an automated integration of weekly preventive maintenance sched ules and work orders with computerize maintenance management tools. 2.4.3 Virtual Reality (VR) A gradual increase of the interest in Virtual Reality (VR) in the past two decades has been witnessed by the Architecture, Engineering, Construction and Facility Management (AEC/FM) industry. VR helped these industries by improving the existing work proces ses. VR is an immersive multimedia technology that creates an enriched virtual environment and allows users to interact with the digital objects in real time (Warwick et.al 1993). VR has been used to tackle a range of design, construction and operation pro blems, including design coordination (Messner et al. 2003, Whyte 2003), project planning (Du et al. 2016), construction education (Kaufmann et al. 2000,

PAGE 22

22 Rekapalli and Martinez 2007), safety training (Sacks et al. 2013), construction operations coordination (Bouchlaghem et al. 2005), facility management (Shi et al. 2016) and real estate (Chen 1995). Due to the fact that BIM includes numerous amounts of information, this makes it a source material for virtual simulations (Macdonald 2012). Evidence indicates a number of advantages of those applications in improving the common understanding among project participants in different phases (Whyte et al. 2000) including: improved design process, reduced misunderstanding, better construction hazards recognition and safety awareness (Dawood et al. 2014, Lin et al. 2011, Miller et al. 2012, Shiratuddin and Thabet 2011). Attributed to the latest technological development, potential VR implementations to transform project communication paradigm is achievable because a BI M based game engine has been extended to VR. VR can be defined as a cybernetic immersive setting that players can adjust and manipulate in real time (Warwick et al. 1993). The difference between the traditional BIM based game engines and VR is that VR not only provides the interactions with various construction components (Kamat et al. 2011), but also provides an immersive experience to the participants (Biocca and Delaney1995). Cumulative evidence has indicated that VR can provide a powerful illusion of pr esence (Hoffman et al. 2003) and triggers similar user behaviors as in physical environments (Heydarian et al. 2015). As a result, VR is recognized as a promising method to boost the standard and quality of the entire AEC/FM workflow, (Messner et al.2003, Park and Kim 2013, Whyte 2003). 2.4.4 Tunneling Design BIM is a broadly adopted concept for the design, construction, and management of buildings or industrial facilities over their whole lifecycle. The development of a

PAGE 23

23 AutodeskTM Revit plugin that was imp lemented with Dynamo was capable of automatizing a design through analysis workflow solution for segmented tunnel lining (Ninic et al. 2019). Recently, large infrastructure projects, such as tunneling, have been increasingly adopting BIM for tunnel design and construction management (Smith 2014, Daller et al. 2016). The procedure is not complex but rather systematic and requires a few parameters that can be easily and directly extracted from the design model. The framework enables the investigation of tunne l alignment alternatives for very long sections in a user friendly and computationally efficient way, thus reducing the manual modeling work substantially (Ninic et al. 2019). 2.4.5 Automatic Generation of Path Networks for Evacuation Another plugin availa ble for AutodeskTM Revit is the Evacuation Path Network (EPN), which is the foundation of route planning, is a network or pathway generated by different available paths in buildings. Manually constructing a graph network or pathways is slow and laborious since buildings are complex and a system cannot be standardized or generalized because each building is different. The approaches used in previous studies contained mainly the visibility graph (VG), generalized Voronoi graphs (GVG), and straight skeletons (Lin and Lin 2018). The end points of obstructions are used by VG as nodes and then connects any two nodes with beelines. Its drawback is that the complexity of the network will increase sharply with the increasing number of nodes (Lin and Chu 2016). GVG i s intended to define the boundary of a generalized Voronoi diagram by means of straight lines. A later study proposed an approach of generating indoor path networks automatically in the BIM platform as a plugin. This new approach has 2 advantages: 1) Since the generation of the path is directly in the BIM platform, there would not be any critical building information loss due to data or

PAGE 24

24 information exchange with another software. 2) Path networks are valuable for building design and can help improve it. Des igners can use path networks to analyze and optimize building layout, such as calculating and identifying the best location of different aspects of the buildings such as identifying the best location of exits for evacuation were all room will be taken into consideration and the optimum position will be identified (Fu and Liu 2019). 2.5 Chatbots in Different I ndustries Robots, artificial intelligence and service automation (RAISA) are already part of real life and not just movies and science (Agah et al. 201 6, Ferreira et al. 2017). After the advancement in RAISA development (Neapolitan and Jiang 2013) and the several economic sectors use RAISA to make their operation proce sses better, improve their costs, enlarge their service capacity, and make customer experience better. RAISA are frequently utilized in education (Timms 2016), medicine (Mirheydar and Parsons, 2012, Schommer et al. 2017), agriculture (Driessen and Heutinck 2014), warehousing and logistics (Min 2009), financial trade (Dunis et al. 2017), manufacturing (Colestock 2005, Pires 2011), transportation (Maurer et al. 2016), journalism (Bollier 2017, Clerwall e of online communication is through chatbots rather than human beings (Hill et al. 2015, Xu et al. 2017). RAISA have already been implemented in travel, tourism and hospitality, though with dissimilar achievement in the numerous tourism sectors (Ivanov e t al. 2017, Kuo et al. 2017, Murphy et al. 2017). Examples can now be seen at numerous airports: online check in, self check in booths, electronic boarding passes, and robotic border control gates with biometric ID card and passport readers. Other examples are F&B vending machines all

PAGE 25

25 over the world, an all automated hotel in Japan, and conveyor belt restaurants in big cities (NY Daily News 2020). 2.5.1 Education The number of students per professor has been continuously increasing in the past decades. (Ni col and Macfarlane 2006). The typical learning scenario is huge lectures at college with more than a 100 student per class and massive open online courses (MOOCs). As a result, personalized support from professors is almost impossible and students are not able to take part in active effective learning (Brinton et al. 2015). Many studies showed that not having such personalized support made weaker learning outcomes and high dropout rates and discontent (Brinton et al. 2015, Eom et al. 2006, Hone and El Said 2016). The ideal solution would be to have one teacher for every student, but clearly this is not a feasible solution due to many fiscal and organizational limitations (Oeste et al. 2015). One promising tool that showed significant positive impact on stude nt satisfaction and learning success is the Chatbot. We can see the effective implementation of chatbots in learning scenarios according to a few studies (Kerly et al. 2007). One example is the University of Georgia, that created created to handle forum posts by students who registered for computer science classes (Goel and Polepeddi 2016). Because of that, scholars became more engaged in the classes and even wanted the same pr ospect for their other courses (Lip ko 2018). Chatbots have a big chance of compensating the lack of individualized support from professors, especially in large scale learning scenarios at universities or in massive open online courses (MOOCs) (Hone and El Said 2016). To elaborate further, chatbots can aid in giving the maximum amount of personalized learning support even with

PAGE 26

26 minimal financial and organizational resources of universities. Chatbots play a huge role in management education and not just the g eneral education benefits cited. Decisive competencies for several managerial competency frameworks are judgment and decision making, feedback providing and receiving, analytical thinking and technological awareness. (Ruth 2006). Yet again, chatbots can de finitely help evolve each of the mention skills. First, chatbots can accurately convey future managers the correct information at the exact time in order to make correct judgments and decisions. Second, chatbots are capable of giving continuous feedback to professors and students. Third, analysis of relevant data by enabling students to analyze a problem first and then receiving the information quickly. Finally, future managers can gain experience in working closely with digital assistants, which will be the norm in future management activities. 2.5.2 Medical Chatbots simplify the communication with electronic health records and even lessen the quantity of physical pape r documents for doctors and physicians. Doctors type out or dictate consultation notes to create modern electronic health records, but this leads to burnouts, loss of information, cognitive load, and a type of distraction from other essential clinical work . Chatbots can also be used to rapidly and easily obtain information about drug interactions and side effects. For example, Safedrugbot is a chat messaging service that aids health professionals in gaining access to information about whether a certain drug is safe for breastfeeding women. There are also chatbots for patients to book appointments and do other tasks for customers. Also, chatbots are turning into the first contact point for many people in terms of identifying symptoms of

PAGE 27

27 certain sicknesses or even recommending what to do next. Although such chatbots are not made to give diagnoses, they can guide patients and even help identify if a serious illness is present or if one should go to a doctor. Some other chatbots are built to th status and even help in managing chronic illnesses. For example, the chatbot Florence works as a personal nurse: it gives reminders to take medicines, gives future instructions in case someone forgot to take the medicine, helps in motivating the user in reaching his/her health targets and delivers medical information (Bates 2019). 2.5.3 Tourism Internet and phones have completely changed the way tourists and travelers plan and experience their trips. The most demanded professions and the way travel destinations advertise themselves has been affected by this this advanced technology. In recent years, using Internet of Things (IoT), VR, AR, and AI to name a few, have be en the most important travel technologies trends. From all the great technologies that are emerging and influencing the travel industry, on big trend is AI and specifically Machine Learning. Traveling is a personal preference: where one wants to vacation, how long they choose to stay, which airlines to travel, and even what to eat are all individual preferences. (Duinkerken 2018) Planning for a trip involves travelers checking for reviews from different websites, collecting correct and up to the minute info rmation, and selecting the most suitable itinerary for their trip. The later proves how essential it is to have personalized and efficient solutions for travel planning. AI, making personalized, programmed, and smart travel services, allows for less challe nging travel planning than previous times (NewGenApps 2018). In detail, Al and Machine Learning are capable of

PAGE 28

28 learning behaviors, choices, and preferences of tourists and thus, can provide a tailer travel experience (Duinkerken 2018, NewGenApps 2018). Cha tbot is a perfect example of Machine Learning because it is capable of creating conversations with humans. Right now, many companies have been employing Chatbots in their products and services like Expedia, KLM, Booking.com, etc (Marques 2018). With the pr ogression of voice interface, chatbots can even communicate with humans using voice technology, like Alexa by Amazon, Cortana by Microsoft, Google Assistant, and Siri by Apple (Huston 2017). Arnold (2018), found that millennials preferred using chatbots an d self service choices rather than communicating with real people. In addition, chatbots can play a crucial role in offering millennials the preferred user experience when organizing their next holiday. 2.5.4 Journalism News corporations such as the BBC ar e acknowledging the divided audiences and the increasing competition from mobile, social, and digital media (Picard 2010, Anderson 2012). Both commercial and public news corporations are under pressure to reach new and underserved audiences whether their p urpose is commercial (increasing advertising revenue) or democratic (better informed citizenry and a healthy public sphere). Just like many PSM, the BBC is confident that the systems of digital innovation will appeal to the younger audiences whose changing consumption patterns show a move to online and on demand viewing (van Es 2017) and social media platforms. Sehl et al.(2016) identified three fundamental interconnected challenges for PSM: Shifting to personal and mobile media, keeping the younger audienc es, and creating effective ways of conveying public service news through third party platforms such as social media, messaging apps, search engines, and video hosting sites. Such encounters

PAGE 29

29 drove PSM to post content on off site platforms like social media and messaging apps. by means of CUIs intended to make the news more attractive and suitable to these methods that has a goal of achieving this kind of journalism by engaging pe ople who are not usually news type of users through the leveraging of new technologies to make layouts that are new, casual, and interactive. The BBC is aspiring to attract a younger at a less formal Labs 2018). The drive of creating new types of conversation with news audiences goes way before CUIs and is rooted in wider changes to the relationship be tween journalism and audiences in the age of the internet where the digital technologies are the center of to many broadcasting system to a many to 1, 532). News bots are also growing within larger economic processes influencing the journalism industry. Technologies like bots, which assure proficiency savings can exhibit augmented engagement of young audiences, are ideal for public news organizations with reduced budgets and they also put to rest the fears of the declining importance of future generations. Bots also represent changes in newsroom culture, especially towards personalization (Helberger 2015) and the growing measurement and tracking of aud iences (Carlson 2018).

PAGE 30

30 CHAPTER 3 RESEARCH METHODOLOGY This section and as highlighted in figure 3 1 describes the rational logic and workflow for creating AutodeskTM Revit add ins tools and then discusses the process of development and evaluation. Two chatbots were created each with a different function, one was created using visual studio and C# while the other was developed using IBM Watson. One bot add in, guides and navigates the user to locate the correct tool to use er place a column and where this tool is located. Another bot add in will be able to handle repetitive tasks such as locating an element, identifying the properties of a c ertain element and quantity take correct value. The chatbots developed throughout this study are aime3d at transforming AutodeskTM Revit to user friendly and more intuitive software by guiding and informing users about what tool and the location of the tool to be used or helping them out with any required calculation in a faster and more effective way leading t o an increase in productivity. These chatbots consequently make construction companies value AutodeskTM Revit more and be considered by a larger number of companies. The following is a list of required resources and skills: AutodeskTM Revit API Platform,20 14 A working understanding of AutodeskTM Revit (Architectural, Structural, MEP). Familiarity with a Common Language Specification compliant language like C#. Microsoft Visual Studio. Microsoft .NET Framework 4.0. The AutodeskTM Revit Software Developer' s Kit (SDK) IBM Watson Assistant

PAGE 31

31 IBM Cloud Figure 3 1 . Rational Logic and Workflow 3.1 Chatbot #1 (Query Chatbot) AutodeskTM Revit API works with the .NET framework, a software framework developed by Microsoft that runs primarily on Microsoft Windows and provides language interoperability. There are various types of programming languages working with the AutodeskTM Revit API, s uch as C#, VB.NET, Ruby, and Python. The most widely used programming language for these applications is C# (Parsons 2010). The majority of SDK samples are in C#, and most online examples are in C#. Even the recommended programming language of Autodesk com pany is C#. The add in or plugin can be described as a software component that is developed to add or modify specific features to an existing computer program. When creating a AutodeskTM Revit add in, or any .NET application, a complier that can convert pr ogram code and compile it into an EXE or DLL file is required. One such compiler for C# is Visual Studio. In a basic program package, it may contain a single EXE file and may be accompanied with one or more DLL files. EXE means an executable program. An EX E file contains an entry point or a similar part in the code where the operating system is supposed to begin the execution of the program, which indicates EXE could be executed on its own.

PAGE 32

32 DLL stands for Dynamic Link Library which is a group of software th at is made available and ready for programs to use (Notenboom 2012). That means that when running a program, such as an EXE file, the program might load additional DLLs to make up the program. For example, in the Revit.exe case, the add in manager plugin i s placed in a separate DLL file that Revit.exe loads every time at startup or when using this plugin. Every DLL corresponds to a function in Revit.exe. Generally, a software is broken up into or provided as an EXE and a collection of DLLs. Launching an EXE means creating a process for it to run on and a memory space. This is necessary for the program to run properly (Joan 2011). The DLL does not have its own memory space and process and is always launched by another application. It simply shares the process and memory space with the application that is calling it. Such structure executes an EXE program with the minimum memory space and maximum function. 3.1.1 Creating Revit Plugin The software used to develop a Revit plugin is the Microsoft Visual Studio. Th ree subscription plans for visual studio are available: Community, Professional and Enterprise. For the development of this project the Visual Studio Community plan was picked since this plan includes all the features that are needed and is free to downloa d and use. Steps to c reate the p lug in 1. Step 1: Create a new Project Using Visual Studio 2. Step 2: Choose Class Library ( .NET Standard) 3. Step 3: Configure your new project. Figure 3 2 shows an example on how to configure a new project. The Following Informatio n are needed:

PAGE 33

33 Project name Location on your Pc Figure 3 2 . Project Set up and Configuration 4. Step 4: Adding References to your Project Referencing the AutodeskTM contents. There are two files that need to be reference: RevitAPI.dll and RevitAPIUI.dll. Adding the references allows and provides access to the Revit database. Without this access, Revit objects and elements cannot be referred to or used as part of the chatbot. RevitAPI.dll: This file provides access to AutodeskTM Revit at a database level which includes all the classes and functions required to create and work with Revit elements. RevitAPIUI.dll: Thi s file provides access to the AutodeskTM Revit user interface which includes all the classes and functions required to add interface elements. These files come with the AutodeskTM Revit install and are compiled codes. The .dll extension stands for Dynamic Link Library. Figure 3 3 shows how the Revit API dll file and the Revit API UI dll file are imported and added to the project . Referencing these files allows access to the required database that enables the creation of the chatbot and commands.

PAGE 34

34 Figure 3 3 . Adding References to the Project After the two referenced files are added and since AutodeskTM Revit already has the AutodeskTM Revit API in it, Copy Local of both files should be set to no or false to ensure they are not copi ed to the output folder compiling the code and when the codes are executed in AutodeskTM Revit. In order to access the contents of the namespace or 4) should be used as a directive that guides the C# compiler to access the required contents. Figure 3 4 . Adding Namespaces to access required contents 5. Step 5: Implement IExternalCommand and IExternalApplication class is accessible from other classes in this plug

PAGE 35

35 (IExternalApplication) and (IExternalCommand) are the 2 types of interfaces for AutodeskTM Revit add ins. An interface can be de scribed as a blueprint for classes that provides a set of methods that a class needs to implement or have in the class. Interfaces define what the class provides to an application. The creator defines the only function the IExternalCommand interface has. The interface that has the Execute function has three variables: CommandData of ExternalCommandData type, message of string type, and finally Elements of ElementSet collection type. ExternalCommandData: Th is object contains reference to the AutodeskTM Revit Application and view required by the external command. All AutodeskTM Revit data can therefore be accessed through this object. ref string: This can be used to relay a message back to the user if the com mand fails or is cancelled. If the command is cancelled or fails, and the string parameter has been set in the command will simply exit. The ref keyword simply marks the string as a reference, meaning that updating it inside of the Execute method, updates a string variable outside of the method scope, stored by AutodeskTM Revit. If it is updated inside of the method, it will update the variable stored by AutodeskTM Re vit. ElementSet: This acts as a list of AutodeskTM Revit elements and can be used to display elements back to the user. If the command fails, and elements have been added to the ElementSet in the command, these will be highlighted to the user. The execute method does not necessarily need to make use of the string and element set parameters however, these are useful for relaying information back to the user. The Execute method needs to Return a result, as marked in the method declaration. This is

PAGE 36

36 a type of e numeration which is either Succeeded, cancelled or failed. Succeeded means the command executed successfully, cancelled means the user cancelled the command and any string set to the input string parameter is sent back to the user. Failed means there was a n unhandled error in the command and any string and ElementSet set to the input parameters is returned to the user. The second type of interface is the IExternalApplication which has two functions: OnShutdown and OnStartup. A single parameter called appli cation is passed in the functions OnShutdown and OnStartup. The application parameter is UIControlledApplication type, which signals back a reference to Revit UI. Revit UI permits consumers to interact with a group of directions prearranged so that the mai n menu is distributed into several command boxes called RibbonTab and each RibbonTab groups several RibbonPanel. RibbonItems are the base case of a menu item in Revit. Each menu item must be associated with a class that implements the interface IExternalCo mmand in order to execute a function (Jesus et al. 2015). After implementing the interface, the process of writing the code starts. Figure 3 5 shows an example of the code script. Figure 3 5 . Code Script Example

PAGE 37

37 In this chatbot development both interfaces were used. The IExternalCommand was used in order to retrieve elements from the Revit database such as counting how many windows are in a project. The IExternalApplication was used in order to create a Text box i n a separate panel for the chatbot were the user writes a question and presses enter to receive an answer. 6. Step 6: Manifest File After writing the code for the needed command, the next step is to register the command for it to show up in AutodeskTM Revit, this is done by creating a manifest file (Figure 3 6). Once AutodeskTM Revit launches, it will scan manifest files found in one of two precise files in order to establish what plugins load and with what options. Application Manifest file is picked from the list of items available to add then renamed to the same name as the plugin, so Revit is aware that they are linked. Also, the extension of the file is changed to .addin because Revit necessitates this type of file. This type of file is called XML file a language formed of tags or markups. Figure 3 6 . Manifest File Script Example The addin file now contains a series of tags that AutodeskTM Revit will read when the plugin starts up. These are not the only tags that can be added, there are many more howe ver, these will provide the information AutodeskTM Revit needs:

PAGE 38

38 in being loaded is either a command or an application . In that case both types are needed and are added separately. to this manifest file. This is so Revit knows which file needs to be loaded. This can be a full file path to a separate file. No file path indicates the dll file is in the same location as this manifest file. 1AB3 4a4b B09F 8C15DFEC6BF0: A GUID which needs to be unique for each command/application. 1: The class that this command/application relates to including the namespace. in Revit. This does not apply to the application. ion of the what the code does: The description that will appear with the command in Revit. This does not apply to the application. 7. Step 7: Test the command and application created After completing the code and manifest file, the DLL file that AutodeskTM Revit can run is created. Testing the code is best done in Debug mode which will highlight any errors that are identified in the code created. To do that, some options need to be cha nged in the Visual Studio file. Activate the Start External Program option (Figure3 7) and browse to the file path to the Revit.exe file. This is found at C: \ Program Files \ Autodesk \ Revit 2020 \ Revit.exe Figure 3 7 . Activating the Start External

PAGE 39

39 Now Autodesk TM Revit will start up each time the code is debug. For Autodesk TM Revit to load the compiled code, it is needed to ensure they are in the correct location when Autodesk TM Revit loads. One of these files is C: \ Users \ user \ AppData \ Roaming \ Autodesk \ R evit \ Addins \ 2020. A build event helps to automatically transfer the compiled code into this directory. This will copy the dll file that the code compiles to, along with the add in file, into the folder that Autodesk TM Revit checks when booting up. The Plug in will be located in the Add ins Tab on the upper left. This chatbot addin uses the UI Application interface that includes a Ribbon Panel system in the code, therefore the add in is going to be provide as a separate panel. Figure 3 8 shows the location of the write the question. After the user writes the question and presses Enter the answer will be shown on a pop up window. Figure 3 8 . Testing the Program 3.1.2 Query Chatbot Experiment For the Query chatbot 36 participants were asked to do a specific task related to Quantity take off on AutodeskTM Revit while using the Schedule of Quantity or Material

PAGE 40

40 that is provided in AutodeskTM Revit once and another run while usin g the developed chatbot. Step 1: The experimented person is asked to do a quantity take off for a certain element on AutodeskTM Revit. Step 2: In the first trial/run the participant is asked to find the quantity of a certain element by using the Schedule o f Quantity or Material and the time it took him to find the answer is recorded. Step 3: In the second trial/run the same participant is asked to find the quantity of the same element but with the help of the Query chatbot and the time to find the answer wa s recorded. Step 4: After both trials/runs are recorded, the participant is asked four different questions: o Is the Chatbot Simple? o Is the Chatbot Clean? o Is the Chatbot Intuitive? o Is the Chatbot Reliable? Note: For each question to be Recorded as YES or TRUE, the chatbot should not only be seen as (Simple, Clean, Intuitive and Reliable) but also should be a better option then the Schedule of Quantity or Material option. For example, if the participant sees that the chatbot is Reliable however the AutodeskTM Revit Schedule of Quantity or Material is more Reliable than I would record it as false or give the point to the Schedules. 3.2 Chatbot #2 (Guiding Chatbot) IBM Watson Assistant is a conversation AI pla tform that helps provide users with fast, straightforward and accurate answers to their questions across any application. 3.2.1 Steps to Develop a Chatbot 1. Step 1: Create workspace in Watson conversation service instant. 2. Step 2: Create and choose a skill for the IBM Watson chatbot that will satisfy the intended goal of the chatbot.

PAGE 41

41 The following types of skills are available: Dialog skill: Uses Watson natural language processing and machine learning technologies to understand user questions and reque sts and then responds to them with answers that are authored by the creator. Search skill: For a given user query, uses the IBM Watson Discovery service to search a data source the creator chooses and return an answer. In this case the chatbot will be able to work as a search engine. (This feature is only available for premium and plus members). * The skill that was used to create the Guiding chatbot is the dialog skill 3. Step 3: Train the Conversation service instance to recognize and identify keywords, conce pts or phrases from the user's input (intents and entities). A minimum of five examples are required for minimal training. The more examples provided the more accurate and efficient the results are. Train the Conversation service instance for each possible intent and entity with natural language examples by adding all synonyms that the user is expected to use. The interface permits the refining of this process later by adding more synonyms as you test your dialog. 4. Step 4: Create a workflow of the stages of the dialog like a mapping system were the Chatbot can identify the intents and entities and will map it and activate a node that will either reply with an answer or will collect more input from the user to answer in a more efficient way. Logical conditions are used for evaluating the concepts identified 5. Step 5: Test how the chatbot will interact with the inputs. Watson Conversation service has a system that allows the creator to monitor how the system interprets the dialog flow and to o verlook on how entities and intents are being detected which will allow the creator to adjust any error or any logical error. 3.2.2 Conversation C oncepts Watson Conversation service includes some concepts and keywords that will be described in this secti on. 3.2.2.1 Intents and entities Watson Conversation services that uses a natural language processing (NLP) is trained to identify keywords and concepts from the users. Intents and Entities (Figure 3 9) are two categories of information that the Conversat ion service extracts in terms of user input.

PAGE 42

42 Figure 3 9 . Intents and Entities (adapted from Azraq et al. 2017) answering a question. By recognizing the intent Assistance service chooses the correct dialog flow for responding to it. These categories are trained using representative examples Recognizing the intents is a way to guide the dialog flow in the appropriate direction and does not require knowing the specifics of the user request. Figure 3 10 shows an example of the intent (unit format). Figure 3 10 . IBM Watson Intent Example An entity expresses a class of obj

PAGE 43

43 Assistance service can pick the particular action(s) to take to satisfy an intent. It is also Watson's method of dealing with s ignificant parts of an input that ought to be utilized to adjust the manner in which it reacts or responds to the intent. Figure 3 11 shows an example of an entity. Entities are the subjects of intents and are explicit qualities or values that clarify user tuned actions and responses. For each value, you can incorporate a rundown of equivalent words or synonyms to catch the potential assortments in user expression. Entities represent information in the user input that is applicable Figure 3 11 . IBM Watson Entity Example The dialog component of the Conversation service uses the intents and entities that are information and data to provide a valuable response to each user input. The dialog is the logical flow that decides the responses your bot will give when certain intents and/or entities are distinguished. 3.2.2.2 Dialog The users will unlikely provide all of the required information in one pass. Instead, a conversation flow must be organized. Using the flow concept, the user will be asked in steps several questions that are useful to gather all the necessary input to

PAGE 44

44 prov ide a helpful answer. A dialog is a branching conversation flow that characterizes how your application responds when it perceives the defined intents and entities. It is composed of many different branching dialog nodes. Generate a dialog branch for each intent, to gather any necessary information and make a useful response. Figure 3 12 shows the dialog for a unit format flow, which is composed of the following dialog nodes: A main node: A node were the user ask for help with respect to unit format. A repl y after the unit format question has been identified that asks what type of unit format is needed? A child node: A node were the user specifies what type of unit format is needed and a response will be given accordingly. Figure 3 12 . Dialog Format 3.2.2 .3 Dialog node The dialog is comprised of nodes that determines and establishes steps in the conversation. An interactive conversation with the user is created by dialog nodes that are chained together in a tree structure. Each node commences with one or more lines that the chatbot shows to the user to request an information or response. Each node includes conditions that only if met will activate the node and also an output object that defines the response provided. One can think of the node as an if then system(Figure 3 13): if this condition is true, the node is activated, and the corresponding response will

PAGE 45

45 be returned. The simplest condition is when the response is returned with a sole intent es are dialog nodes that do not depend on other nodes. The dialog component of the Assistance service uses ultimately provide a useful response. The dialog is a tree like structure represented as branch to process each intent that is defined. The dialog node is made of a trigger and a response, where the trigger is the condition. Dialog maps intents (what users say) to responses (what the bot says back). Each dialog no de contains at least one condition and one response, and the dialog response determines how to reply to the user. Figure 3 13 . Dialog Concept (adapted from Azraq et al. 2017 ) 3.2.2.4 Context a s in a real life conversation. The process of passing informat ion between the dialog and the application code is called the dialog context. Information are allowed to pass back and forth across different dialog nodes due to the fact that context allows to store information. For example, if the name of the user is det ermined in the Conversation flow, it could be stored in the context and retrieved at any time the username needs to be recalled.

PAGE 46

46 3.2.2.5 Condition and responses Conditions can be described as logical expressions that are evaluated to true or false. Conditions are used to map and select the next dialog node in the flow, or to pick among the possible responses to the user. Conditions not only evaluate the intents and entities identified in the user responses but also can evaluate information stored in the context. Either previous dialog nodes or the application code as part of an API call can store information from the context. After the bot identifies and evaluate the intents and entities and the mapping node is activated a response is communicated to the user. The creator can add variations of the response for a more natural experience where a number of answers are generated, and the bot chooses randomly between them each time or add conditions within the same dialog to pick one response out of many. 3.2.2.6 Conversation turn A conversation turn (Figure 3 14) is described as a single cycle of user input and a response. Each conversation turn starts with one active node and ends with displaying the response or action of this active node . Figure 3 14 . Conversation Turn (adapted from Azraq et al. 2017)

PAGE 47

47 3.2.2.7 Typical conversation flow 1. The conversation begins with an initial node set up with the conversation_start special condition. 2. After some conversation turns, the bot identifies and examines the user's input that includes the intents and entities then the dialog progresses to the node that is mapped as an active node to these specific intents and entities. The response that is configured in this node is then showed either directly or more step conditions should be answered to show the answer. 3. nodes will be evaluated in descending order. The next active node is selected when the first child node m atches the conditions and a new conversation turn starts. 4. The Conversation service evaluates the conditions of each base node in the dialog and selects the first matching dialog node as the next active node If no child node matches the condition and was ma pped. 5. If no other nodes match the conditions the conversation defaults to a base node configured with the anything_else special condition. The special anything_else not c omp rehended and determined by the bot. In this case the answer can be such 15 shows a typical conversation flow step. Figure 3 15 . Conversation Flow (adapted from Azraq et al. 2017) 3.2. 3 Guiding Chatbot Experiment For the guiding chatbot the 67 participants were asked to find a specific tool in AutodeskTM Revit while using the AutodeskTM Revit search once and another run while using the developed chatbot.

PAGE 48

48 1. Step 1: The participant is asked to find a certain tool in AutodeskTM Revit. 2. Step 2: In the first trial/run the participant is asked to locate the tool with the help of the AutodeskTM Revit Search and the time it took them to locate or find the answer is recorded. 3. Step 3: In the second tr ial/run the same participant is asked to locate the same tool but with the help of the Guiding chatbot and the time to find the answer was recorded. 4. Step 4: After both trials/runs are recorded, the participant is asked four different questions: o Is the Chat bot Simple? o Is the Chatbot Clean? o Is the Chatbot Intuitive? o Is the Chatbot Reliable? Note: For each question to be Recorded as YES or TRUE, the chatbot should not only be seen as (Simple, Clean, Intuitive and Reliable) but also should be a better option th en the AutodeskTM Revit search. For example, if the participant sees that the chatbot is clean however the AutodeskTM Revit search is cleaner than I would record it as false or give the point to the AutodeskTM Revit search. 3.3 Conclusion Intents and enti ties are highlighted and identified from the user's input through the conversation service. This information and context information will define the flow of the conversation or flow of dialog nodes called a dialog. The nodes will be activated depending on its designed and configured conditions and will have a response to show to the user. These straightforward essential concepts permit the creation of a complex, powerful, and practical user interaction experience.

PAGE 49

49 CHAPTER 4 RESULTS AND ANALYSIS 4.1 Experimental Metrics Knowing the great power (capabilities) of AutodeskTM Revit and its diverse aspects in the project design and construction phases, one should note that in order for a software to be implemented in a company it should first be User friendly, it should be described as easy to use. It is "friendly" to the user, meaning it is not difficult, complex or challenging to learn or understand. While "user friendly" is a subjective term, the following are several common attributes found in user friendly interfaces (Christensson, 2014): 1. Simple: A user friendly interface should not be perplexing or complex, but instead is clear and straightforward, providing simple access to most features or commands. 2. Clean: A good user interface is well organized, making it simple to find and locate various tools and options. 3. Intuitive: In order to be user friendly, an interface must make sense to the average user and should require minimal explanation for how to use it, giving users brisk access to regular highlights or directions 4. Reliable: An unreliable product will cause undue frustration for the user. A user friendly product is reliable and does not breakdown, malfunction or crash . Another Important factor that should be taken into consideration is the speed of the reaction or response of the developed chatbots. The objective of a user friendly product is to provide a decent user experience. This may appear to be unique relying up on the end client for whom the item is designed. For instance, a user friendly kid's game will have a much different interface than a professional CAD program. Nonetheless, the principles above apply to both types of software . Regardless of whether a program has many propelled highlights and advanced features, it is yet conceivable to make it easy to understand by planning and

PAGE 50

50 designing a bas ic, clean, and intuitive interface. User friendly products are typically more successful and effective than those with mind boggling, tangled interfaces that are hard to utilize. Easy to understand and straightforward tools are ordinary more productive tha n those with intricate, tangled interfaces that are difficult to use. (Christensson, 2014) 4.2 Data Collection 4.2.1 Data Collection #1: Guiding Chatbot For the Guiding chatbot an experiment was done were 67 different participants were involved. The partic ipants were chosen according to their proficiency level with the AutodeskTM Revit software. Out of the 67 involved people, 28 had Beginner level, 14 had intermediate level and 25 had professional level experience. 4.2.2 Data Collection #2: Query Chatbot F or the Query chatbot an experiment was conducted with 36 different participants. The participants were chosen according to their proficiency level with the AutodeskTM Revit software. Out of the 36 participants, 12 had Beginner level, 12 had intermediate le vel and 12 had professional level experience. 4.3 Results and Analysis 4.3.1 Results and Analysis #1: Guiding Chatbot Figure 4 1 shows the time result and compares the speed difference between the chatbot and AutodeskTM Revit search. As shown in the below graph, the average speed for a beginner level using a chatbot is 6.46 seconds, while the average speed using the Autod eskTM Revit search is N/A as they were not able to find and locate the tool they needed even after a long period of time. With respect to the intermediate levels, the average time that was needed to locate the tool with the help of chatbot was

PAGE 51

51 6.39 seconds compared to 66.03 seconds when using the AutodeskTM Revit search. Lastly, the professionals took 5.76 seconds to locate the tool using the chatbots while it took them 31.13 seconds when using AutodeskTM Revit search. All proffeciency level AutodeskTM Revi t users were able to benift from the chatbot and get a faster answer while using the Guiding chatbot. Figure 4 1 . Guiding Chatbot Speed Results Figure 4 2 shows the results related to the questions each user undertaking the experiment was asked (Simple, Clean, Intuitive and Reliable). 100% of the beginner level users indicated that the chatbot is simple, clean, intuitive and reliable. Also, exactl y all of the Intermediate level users indicated that the chatbot is simple, clean, intuitive and reliable. However, for the expert level users, 100% indicated the chatbot is simple, clean and intuitive but only 75% indicated that it is reliable. None of th e beginners or intermediates indicated that AutodeskTM revit search is simple, clean, intuitive, or

PAGE 52

52 reliable but 25% of the expert highlighted that the AutodeskTM Revit search is more reliable. Figure 4 2 . Guiding Chatbot Experimental Results Referring to the results, all users in the experiment from a AutodeskTM Revit proficiency level of beginner to an expert were able to find and locate the tool needed faster when using the chatbot. The beginner level users were not able to find and locate any tool wi thout the help of the chatbot and the intermediate level users took a lot of time to locate and find a needed tool with the help of AutodeskTM Revit search. This indicates the importance of the chatbot with respect to new and inexpert Revit users. In order for companies to start implementing the Revit software into their daily tasks, some sort of help should be available for the users especially since most if not all of them are going to be of a beginner or intermediate level. The chatbots will allow the smooth transition from other types of software such as AutodeskTM AutoCAD to AutodeskTM Revit. With the help of the chatbots, users will have faster answers on what tool to use

PAGE 53

53 and where it is located, hence leading to a better and efficient productivity. During this study, the effect of the chatbot on expert users was also taken into consideration. Expert users were also satisfied by the chatbot and were able to get a faster answer when using a chatbot. During the experiment, all the beginner, intermediate and expert level users where asked if the chatbot is simple, clean, intuitive and reliable. Not surprisingly, 100% of the beginner and intermediate level users highlighted that the chatbot was cleaner and more organized than the AutodeskTM Revit search, s impler and easier than the AutodeskTM Revit search, and intuitive and can be used without explanation. With respect to the Reliable factor, all users indicated that the chatbot was more reliable except for 25% of the expert users that indicated the Autodes kTM Revit search was more reliable. Those 25% claimed that the Revit search gave them multiple answers or options to choose from that they find more useful. This fact is true but not with respect to beginner and intermediate levels since multiple options w ill just confuse them and make them guess what the right option is to use. Studying the effect of the chatbot on all proffecincy level is important to make sure that the chatbot will benefit every user and if not benifiting them atleast doesnot have any ne gative side effects. 4.3.2 Results and Analysis #2: Query Chatbot Figure 4 3 shows the Time result and compares the speed difference between the chatbot and the Schedule of Quantity or Material. The average speed for a beginner level using a chatbot is 5.7 3 seconds, while the average speed using the AutodeskTM Revit schedule of quantity or material is N/A as the participants were not able to find the answer they needed even after a long period of time. With respect to the intermediate levels, the average ti me that was needed to find the answer with the help of chatbot was 5.57 seconds compared to 98.40 seconds when using the AutodeskTM Revit Schedule

PAGE 54

54 of Quantity or Material. Lastly, the professionals took 5.06 seconds to find the answer lp while it took them 22.20 seconds when using AutodeskTM Revit Schedule of Quantity or Material. All proffeciency level AutodeskTM Revit users were able to benift from the chatbot and get a faster and accurate answer while using the Quey chatbot. Figur e 4 3 . Query Chatbot Speed Results Figure 4 4 shows the results related to the questions each user undertaking the experiment was asked (Simple, Clean, Intuitive and Reliable). 100% of the beginner level users indicated that the chatbot is simple, clean, intuitive and reliable. Also, 100% o f the Intermediate level users indicated that the chatbot is simple, clean, intuitive and reliable. However, for the expert level users, 100% indicated the chatbot is simple, clean and intuitive but only 66.67% indicated that it is reliable. None of the be ginners or intermediates indicated that the AutodeskTM Revit Schedule of Quantity or Material is simple, clean, intuitive, or reliable but 33.33% of the experts highlighted that the AutodeskTM Revit search is more reliable.

PAGE 55

55 Figure 4 4 . Query Chatbot Exp erimental Results Referring to the results, all users in the experiment from a AutodeskTM Revit proficiency level of beginner to an expert were able to find the answer in a faster way when using the chatbot. The beginner level users were not able to find t he answer without the help of the chatbot since they did not know how to do a Schedule of Quantity or Material and the intermediate level users took a lot of time to find the answer while using the Schedule of Quantity or Material since they needed time to remember where it was located and how to do the steps. The later indicates the importance of the chatbot with respect to new and inexperienced AutodeskTM Revit users. For companies to start implementing the AutodeskTM Revit software into their daily tasks , some sort of help should be available for the users especially since most if not all of them are going to be of a beginner or intermediate level. The chatbots will allow the smooth transition from other types of software such as AutodeskTM AutoCAD to A utodeskTM Revit. With the help of the chatbots, users will be able to have answers even with null experience with AutodeskTM Revit software. This study also showed that the expert users were getting faster answers while using the chatbot and users claimed that this helped with

PAGE 56

56 the productivity of their work. During the experiment, all the beginner, intermediate and expert level users where asked if the chatbot is simple, clean, intuitive and reliable. 100% of the beginner, intermediate and experts level use rs highlighted that the chatbot was cleaner and more organized than the AutodeskTM Revit Schedule of Quantity or Material, simpler and easier than the AutodeskTM Revit Schedule of Quantity, more intuitive and can be used without explanation compared to Sch edule of Quantity or Material. With respect to the Reliable factor, all users indicated that the chatbot was more reliable except for 33.33% of the expert users that indicated the AutodeskTM Revit Schedule of Quantity or Material was more reliable. Those 3 3.33% claimed that the Revit Schedule of Quantity or Material gave them an answer that includes many factors such as length, level and type. For experts this could be an advantage but still they can use the chatbot to get the length, type or any property t hat is needed. However, this chatbot was created and made capable of counting objects for quantity takeoff but could be developed into identifying the type of a certain element, area of a Studying the effect of the chatbot on all proffecincy le vel is important to make sure that the chatbot will benefit every user and if not benifiting them atleast doesnot have any negative side effects.

PAGE 57

57 CHAPTER 5 CONCLUSION AND RECOMMENDATIONS 5.1 Conclusion With respect to the architecture, engineerin g and construction (AEC) industry, BIM is one of the most promising yet challenging development. The main feature of BIM is that it can construct a virtual model of a building or structure. Different aspects of the AEC industry such as designing, planning, construction and facility operation are using and adopting the BIM model, and this concept helps visualize what is to be built in a simulated environment to identify any potential design, construction, or operation issues. With that being said, many compa nies are not adopting BIM software due to the fact that companies are not willing to change from one type of software to another. For examples, design companies with architects that use AutodeskTM AutoCAD and have been using it for decades are going to find it very challenging to shift to AutodeskTM Revit even though AutodeskTM Revit is way more developed. The later can be broken down into two fundamental reasons: First, human nature by itself resists to change and second, AutodeskTM Revit is hard to le arn and is going to demand a lot of training. The guiding chatbot created is developed with an aim to help the users know how to use AutodeskTM Revit and is a type of guidance, which in itself is a training process for the users. Moreover, the Query chatbo t is created with an aim to help the user achieve the intended goal faster, more accurately and more efficiently. Both chatbots will help different level of Revit proficiency achieve the intended goal. Therefore, developing the chatbots will ease the shift of companies from one software to another.

PAGE 58

58 5.2 Recommendations The chatbots (Guiding Chatbot and Query Chatbot) are both a subject to future studies. The Guiding chatbot can be developed to answer a wider range of questions and have multiple answers for any question, as well as it could be further developed to assist all BIM softw ares and not only Revit. The Query chatbot could be re designed with AI in order to give the right answer even if the users misspelled a word or wrote a different keyword. If developed thouroughly, both chatbots could help ease the shift towards BIM softwa res. In addition to that, both chatbots could also be developed into Voice enabled chatbots where the application will no longer be a textbox by its self where the user asks a question in the form of a text but also the user will be allowed to record a question thus allowing the bot to answer accordingly. A user interacts with a voice enabled chatbot differently: they interact with such a bot via their voice in natural language. The voice chatbot then answers back using pre recorded messages, text to sp eech responses or a mixture of both.

PAGE 59

59 APPENDIX CODES SCRIPT Windows Quantity Takeoff Code: using System; using System.Collections.Generic; using System.Linq; using System.Text; using System.Threading.Tasks; using Autodesk.Revit.Attributes; using Autodesk.Revit.DB; using Autodesk.Revit.UI; namespace Chatbot { [TransactionAttribute(TransactionMode.Manual)] public class WindowsCode : IExternalCommand { public Result Execute(ExternalCommandData comma ndData, ref string message, ElementSet elements) { UIDocument uidoc = commandData.Application.ActiveUIDocument; Document doc = uidoc.Document; FilteredElementCollector collector = new FilteredElementCollector(do c); ElementCategoryFilter filter = new ElementCategoryFilter(BuiltInCategory.OST_Windows);

PAGE 60

60 IList windows = collector.WherePasses(filter).WhereElementIsNotElementType().ToElements(); TaskDialog.Show("Windows", windows.Count + "windows found"); return Result.Succeeded; } } } Doors Quantity Takeoff Code: using System; using System.Collections.Generic; using System.Linq; using System.Text; using System.Threading.Ta sks; using Autodesk.Revit.Attributes; using Autodesk.Revit.DB; using Autodesk.Revit.UI; namespace Chatbot { [TransactionAttribute(TransactionMode.Manual)] public class DoorsCode : IExternalCommand { public Result Execute(ExternalCommand Data commandData, ref string message, ElementSet elements) {

PAGE 61

61 UIDocument uidoc = commandData.Application.ActiveUIDocument; Document doc = uidoc.Document; FilteredElementCollector collector = new FilteredElementCol lector(doc); ElementCategoryFilter filter = new ElementCategoryFilter(BuiltInCategory.OST_Doors); IList doors = collector.WherePasses(filter).WhereElementIsNotElementType().ToElements(); TaskDialog.Show("Windows", doors.Count + "doors found"); return Result.Succeeded; } } } Chatbot interaction Code: public class nocom : IExternalApplication { public Result OnShutdown(UIControlledApplication app) { return Result.Succeeded; } public Result OnStartup(UIControlledApplication a pp) { RibbonPanel ribbonPanel = app.CreateRibbonPanel("Chatbot");

PAGE 62

62 TextBoxData textData = new TextBoxData("Text Box"); TextBox tBox = ribbonPanel.AddItem(textData) as TextBox; tBox .PromptText = "Enter A Question"; tBox.ToolTip = "Ask me a question"; tBox.EnterPressed += new EventHandler(ProcessTex t); return Result.Suc ceeded; } void ProcessText(object sender, Autodesk.Revit.UI.Events.TextBoxEnterPressedEventArgs args) { TextBox tBox = sender as TextBox; string strText = tBox.Value as string; if ( strText == ("how many doors")) { var DoorsCode = new DoorsCode(); DoorsCode.Execute(sender as UIApplication); } if (strText == ("how many windows")) { var WindowsCode = new WindowsCode(); WindowsCode.Execute(sender as UIApplication);

PAGE 63

63 } if (strText != ("how many windows")&& strText != ("how many doors")) { TaskDialog.Sh ow("Windows: ", " Ask Another Question"); } } } }

PAGE 64

64 LIST OF REFERENCES Agah, A., Cabibihan, J., Howard, A., Salichs, M., and He, H. (2016). "Social Robotics". Proceedings of the 8th International Conference on Social Robotics (ICSR), 104 114. Akbarnezhad, A., Ong, K., and Chandra, L. (2014). "Economic and environmental assessment of deconstruction strategies using building information modeling". Automation in Construction, 37, 131 144. Anderson, Chris W. 2011. Deliberative, Agonistic, and Algorithmic Audiences: Journalism's Vision of its Public in an Age of Audience International Journal of Communication 5: 529 547. Anorld, A. 2018. Why Millennials Prefer Chatbots Over Travel Agents When Planning Their Vacations. URL: https://www.forbes.com/sites/andrewarnold/2018/02/24/why mille nnials prefer chatbots over travel agents when planning their vacations/#78d584752233 . Azhar, S., and Brown, J. (2009). "BIM for Sustainability Analyses". International Journal of Construction Education and Research, 5(4), 276 292. Azraq, A., Aziz, H., Nap pe, N., and sri, l. (2017). Building cognitive applications with IBM Watson Services: Volume 2 Conversation. Barnes, s. and Castro Lacouture, D. (2009). BIM enabled integrated optimization tool for LEED decisions. International Workshop on Computing in Civ il Engineering. Bates, M. 2019. "Health Care Chatbots Are Here to Help" . IEEE Pulse, 10, 3,12 14. BBC News Labs. Medium. June. https://medium.com/bbc news labs/bbc botbuilder ba8e09b6a2e9 . Biocca, F., & Delaney, B. (1995). Immersive virtual reality technology. Communication in the age of virtual reality (p. 57 124) Bollier, D. (2017). Artificial intelligence comes of age. The promise and challenge of integrating AI into cars, healthcare and journalism. The Aspen Institute, Washington, DC. Bouchlaghem , D., Shang, H., Whyte, J., and Ganah, A. (2005). "Visualisation in architecture, engineering and construction (AEC)". Automation in Construction, 14(3), 287 295. Brinton, C., Rill, R., Ha, S., Chiang, M., Smith, R., and Ju, W. (2015). "Individualization f or Education at Scale: MIIC Design and Preliminary Evaluation". IEEE Transactions on Learning Technologies, 8(1), 136 148.

PAGE 65

65 Carlson, Matt. 2018. Digital Journalism, 6 (4): 406 441 Chen, S. (1995). "QuickTime VR: An Image Based Approach to Virtual Environment Navigation". Proceedings of the 22nd Annual Conference on Computer Graphics and Interactive Techniques, 29 38. Chen, A., and Chu, J. (2016). "TDVRP and BIM Integrated Approach for In Building Emergency Rescue Routing" . Journal of Computing in Civil Engineering, 30(5). Clerwall, C. (2014). "Enter the Robot Journalist". Journalism Practice, 8(5), 519 531. Colestock, H. (2005). Industrial robotics. McGraw Hill, New York, NY. . (2016). "Implementation of BIM in the tunnel design Engineering consultant's aspect". Geomechanics and Tunnelling, 9(6), 674 683. Dawood, N., Miller, G., Patacas, J., and Kassem, M. (2014). "Combining Serious Games and 4D Modelling for Construction Hea lth and Safety Training". Computing in Civil and Building Engineering (ASCE), 2087 2094. Driessen, C., and Heutinck, L. (2014). "Cows desiring to be milked? Milking robots and the co evolution of ethics and technology on Dutch dairy farms". Agriculture and Human Values, 32(1), 3 20. Du, J., Shi, Y., Mei, C., Quarles, J., and Quarles, W. (2016). "Communication by Interaction: A Multiplayer VR Environment for Building Walkthroughs". ASCE Construction Research Congress. Duinkerken, W.v. 2018. How Machine Learn ing is Changing the Travel Industry. URL: https://jaxenter.com/machine learning travel industry 147766.html . Dunis, C., Theofilatos, K., Karathanasopolous, A., and Middlet on, P. (2017). Artificial Intelligence in Financial Markets: Cutting Edge Applications for Risk Management, Portfolio Optimization and Economics. Palgrave Macmillan, London. Elupula, V. (2019). "How do chatbots work? An overview of the architecture of a ch atbot". Big Data Made Simple, (Oct. 20, 2020). Eom, S., Wen, H., and Ashill, N. (2006). "The Determinants of Students' Perceived Learning Outcomes and Satis faction in University Online Education: An Empirical Investigation*". Decision Sciences Journal of Innovative Education, 4(2), 215 235.

PAGE 66

66 Ferreira, M., Sequeira, J., Tokhi, M., Kadar, E., and Virk, G. (2017). "A World with Robots". International Conference o n Robot Ethics: ICRE 2015, 84. Forns Samso, F. (2010). "Perceived value of building information modeling in facilities operations and maintenance". Masters, University of New Mexico. Fu, M., and Liu, R. (2019). "Automatic Generation of Path Networks for Evacuation Using Building Information Modeling". ASCE International Conference on Computing in Civil Engineering 2019. Gallaher, M. P., O'connor, A. C., Dettbarn, J. J., & Gilday, L. T. (2004). Cost Analysis of Inadequate Interoperability in t he U.S. Capital Facilities Industry. U.S. Department of Commerce Technology Administration. Gandhi, S. and Jupp, J. (2014). BIM and Australian Green Star Building Certification. Computing in Civil and Building Engineering, ASCE (2014), pp.275 282. Goel, A. , Polepeddi, L. 2016. Jill Watson: a virtual teaching assistant for online education. Georgia Institute of Technology GSA. (2007). "GSA Building Information Modeling Guide 01 BIM Guide Overview". U.S. General Services Administration, http://www.gsa.gov/gsa/cm_attachments/GSA_DOCUMENT/GSA_BIM_Guide_v 0_60 _Series01_Overview_05 14 07_R2C a3 l_0Z5RDZ i34KpR.pdf . S. Gupta, M. Keen, A. Shah, G. Verdier (2017). Digital revolutions in public finance. International Monetary Fund. Helberger, Natali. 2015. Intern ational Journal of Communication 9: 1324 1340. Heydarian, A., Carneiro, J., Gerber, D., Becerik Gerber, B., Hayes, T., and Wood, W. (2015). "Immersive virtual environments versus physical built environments: A benchmarking study for building design and use r built environment explorations". Automation in Construction, 54, 116 126. Hill, J., Randolph Ford, W. and Farreras, I. (2015). Real conversations with artificial intelligence: A comparison between human human online conversations and human chatbot conver sations. Computers in Human Behavior, 49, pp.245 250. Hoffman, H., Richards, T., Coda, B., Richards, A., and Sharar, S. (2003). "The Illusion of Presence in Immersive Virtual Reality during an fMRI Brain Scan". CyberPsychology & Behavior, 6(2), 127 131. Ho ne, K., and El Said, G. (2016). "Exploring the factors affecting MOOC retention: A survey study". Computers & Education, 98, 157 168.

PAGE 67

67 Hu, J., Wang, Z. and Chen, W. (2020). A Study on Automatic Form Optimization Procedures of Building Performance IOP Conference Series: Earth and Environmental Science, 531, p.012020. us like real people?. URL: https://www.sciencemag.org/news/2017/06/when will alexagoogle assistant and other chatbots finally talk us real people . Inyim, P., Rivera, J., and Zhu, Y. (2015). "Integration of Building Information Modeling and Economic and Environmental Impact Analysis to Support Sustainable Building Design". Journal of Management in Engineering, 31(1). Ivanov, S., Webster, C. & Berezina, K. (2017). Adoption of robots and service automation by tourism and hospitality companies. Paper presented at the INVTUR Conference, 17 19 May 2017, Aveiro, Portugal. Jalaei, F. and Jrade, A. (2014). Integrating Building Information Modeling (BIM) and Energy Analysis Tools with Green Building Certification System, to Conceptually Design Sustainable Buildings, Journal of Information Technology in Construction, ITcon Vol. 19 (2014), 494 519. in design Journal of Building Information Modeling spring 2010., 13 16. Kamat, V., Martinez, J., Fischer, M., Golparvar Fard, M., Peña Mor a, F., and Savarese, S. (2011). "Research in Visualization Techniques for Field Construction". Journal of Construction Engineering and Management, 137(10), 853 862. Kar, R. and Haldar, R. (2016). Applying Chatbots to the Internet of Things: Opportunities a nd Architectural Elements. International Journal of Advanced Computer Science and Applications, 7(11). Kaufmann, H., Schmalstieg, D., and Wagner, M. (2000). "Construct3D: a virtual reality application for mathematics and geometry education". Education and Information Technologies, 5, 263 276. Kerly, A., Hall, P., and Bull, S. (2007). "Bringing chatbots into education: Towards natural language negotiation of open learner models". Knowledge Based Systems, 20(2), 177 185. Klein, L., Li, N. and Becerik Gerber, B. (2012). Imaged based verification of as built documentation of operational buildings. Automation in Construction, 21, pp.161 171.

PAGE 68

68 Koch, C., Vonthron, A., and König, M. (2017). "A tunnel information modelling framework to support management, simulations and visualisations in mechanised tunnelling projects". Automation in Construction, 83, 78 90. Conference. 2010 Krukowski, A., & Arsenijevic, D. (2010). RFID based positioning for building management systems. Proceedings of 2010 IEEE International Symposium on Circuits and Systems. Kumar, S. (2008). " Interoperability between building information models (bim) and energy analysis programs". Graduate, Faculty of the School of Architect ure, University of Southern California. Kuo, C., Chen, L., and Tseng, C. (2017). "Investigating an innovative service with hospitality robots". International Journal of Contemporary Hospitality Management, 29(5), 1305 1321. Lin, W., and Lin, P. (2018). "In telligent generation of indoor topology (i GIT) for human indoor pathfinding based on IFC models and 3D GIS technology". Automation in Construction, 94, 340 359. Lin, K., Son, J., and Rojas, E. (2011). "A pilot study of a 3D game environment for constructi on safety education". Journal of Information Technology in Construction (ITcon), 5, 69 84. Lip ko, H. (2018). Meet Jill Watson: Georgia Tech's first AI teaching assistant; https://pe.gatech.edu/blog/meet jill watson georgia techs first ai teaching assistant Macdonald, J. (2012). " A framework for collaborative bim education across the aec disciplines". 37th Annual Conference of the Australasian Universiti es Building Educators Association (AUBEA) The University of New South Wales, Australia, 4, 4 6. Marques, M. 2018. Top 3 chatbots that are changing the travel industry. URL: https://medium.com/hijiffy/top 3 chatbots that are changing the travel industryd325082c50b8 . Maurer, M., Christian Gerdes, J., Lenz, B., and Winner , H. (2016). Autonomous Driving: Technical, Legal and Social Aspects. Springer, Cham. Messner, J., Yerrapathrun, S., Baratta, A., and Whisker, V. (2003). "Using virtual reality to improve construction engineering education". American Society for Engineerin g Education Annual Conference & Exposition.

PAGE 69

69 Miller, G., Kassem, M., and Dawood, N. (2012). "Building an Emergent Learning Environment for Construction Health and Safety by Merging Serious Games and 4D Planning". International Conference on Computing in Civ il Engineering, 129 136. Min, H. (2009). "Artificial intelligence in supply chain management: theory and applications". International Journal of Logistics Research and Applications, 13(1), 13 39. Mirheydar, H., and Parsons, J. (2012). "Diffusion of robotic s into clinical practice in the United States: process, patient safety, learning curves, and the public health". World Journal of Urology, 31(3), 455 461. Neapolitan, R., and Jiang, X. (2013). "Contemporary artificial intelligence". Boca Raton. NewGenApps 2018. Chatbots in Travel. URL: https://www.newgenapps.com/blog/chatbots in travel . Nicol, D., and Macfarlane Dick, D. (2006). "Formative assessment and self regulated learning: a model and seven principles of good feedback practice". Studies in Higher Education, 31(2), 199 218. Ninic, J., Giang Bui, H., and Meschke, G. (2019). "Parametric Design and Isogeometric Analysis of Tunnel Linings within the Building Information Modelling F ramework". University of Nottingham. "NY Daily News." (2020). Nydailynews.com, . Oeste, S., Lehmann, K., Janson, A., Söllner, M., and Leimeister, J. (2015). "Redesignin g University Large Scale Lectures: How To Activate The Learner". Academy of Management Proceedings, 2015(1), 14650. Park, C., and Kim, H. (2013). "A framework for construction safety management and visualization system". Automation in Construction, 33, 95 103. Pires, J. (2011). Industrial robots programming. Springer, New York. Roudsari, M. and Pak, M. (2013). Ladybug: A parametric environmental plugin for grasshopper to help designers create an environmentally conscious design . 13th Conference of Internati onal Building Performance Simulation Association, Chambéry, France. 3128 3135,. Ruth, D. (2006). "Frameworks of managerial competence: limits, problems and suggestions". Journal of European Industrial Training, 30(3), 206 226.

PAGE 70

70 S, R. and K, B. (2018). Empow ering chatbots with business intelligence by big data integration. International Journal of Advanced Research in Computer Science, 9(1), pp.627 631. Sacks, R., Perlman, A., and Barak, R. (2013). "Construction safety training using immersive virtual realit y". Construction Management and Economics, 31(9), 1005 1017. Schommer, E., Patel, V., Mouraviev, V., Thomas, C., and Thiel, D. (2017). "Diffusion of Robotic Technology Into Urologic Practice has Led to Improved Resident Physician Robotic Skills". Journal o f Surgical Education, 74(1), 55 60. Schwab, K. (2016). The Fourth Industrial Revolution. World Economic Forum, Cologny/Geneva. Schwegler, B. (2010). Green BIM: Successful Sustainable Design With Building Information Modeling by Eddy Krygiel and Bradley Nie s. Journal of Industrial Ecology, 14(5), pp.859 860. Sehl, Annika, Cornia, Alessio, and Nielsen, Rasmus Kleis. 2016. Reuters institute for the Study of Journalism. Shi, Y., Du, J., Lavy, S ., and Zhao, D. (2016). "A Multiuser Shared Virtual Environment for Facility Management". Procedia Engineering, 145, 120 127. Shiratuddin, M., and Thabet, W. (2011). " Utilizing a 3d game engine to develop a virtual design review system ". Journal of Informa tion Technology in Construction, 16, 39 68. Smith, S. (2014). "Building information modelling moving Crossrail, UK, forward". Proceedings of the Institution of Civil Engineers Management, Procurement and Law, 167(3), 141 151. Talwar, R. (2015). "The fu ture of business". Fast Future Publishing. Timms, M. (2016). "Letting Artificial Intelligence in Education Out of the Box: Educational Cobots and Smart Classrooms". International Journal of Artificial Intelligence in Education, 26(2), 701 712. Van Es, Kare n. 2017. Driven Personalization Media@LSE working papers. 43: 18 https://dspace.library.uu.nl/handle/1874/358206 Warwick, K. (2012). "Artificial intelligence: the basics". Oxon: Routledge. Warwick, K., Gray, J., and Roberts, D. (1993). "Virtual reality in engineering". Institution of Electrical Engineers.

PAGE 71

71 Whyte J (2003). Industrial applications of virtual reality in architecture and construction, Journal of Information Technology in Construction, 8, Special issue Virtual Reality Technology in Architecture and Construction, pg. 43 50, Whyte, J., Bouchlaghem, N., Thorpe, A., and McCaffer, R. (2000). "From CAD to virtual reality: modelling approaches, data exchange and interactive 3D building design tools". Automation in Construction, 10(1), 43 55. Wong, J., and Zhou, J. (2015). "Enhancing environmental sustainability over building life cycles through green BIM: A review". Automation in Construction, 57, 156 165. Xu, A., Liu, Z., Guo, Y., Sinha, V., & Akkiraju, R. (2017). A new chatbot for customer service on social media. In Proceedings of the 2017 CHI Conference on Human Factors in Computing Systems,3506 3510 Yan, H., and Damian, P. (2008). Benefits and Barriers of Building Information Modelling. 12th International Conference on Computing in Civil and Building Engineering, Beijing.

PAGE 72

72 BIOGRAPHICAL SKETCH Maroun Waked was born in Lebanon in a family of strong moral values, loving and caring . He started swimming training since he was 6 years old and continued with my training until I became a professional swimmer competing worldwide. In May 2018, he earned h is Bachelor of Engineering in civil engineering and a minor in construction management from the Lebanese American University, Byblos Lebanon. Maroun waked was not satisfied yet with his knowledge about construction so he decided to pursue a M S in construct ion management at University of Florida. H e received his MS from the M.E. Rinker, Sr. School of Construction Management at the University of Florida. Upon education, working towards e arning a Ph.D. degree in construction management from the College of Design, Construction, and Planning at the University of Florida.