A Chatbot System for Construction Daily Report Information Management

Abstract

Consistently updating, analyzing, and managing construction-related information is one of the key success factors in project management. Quite a few construction projects have recently started to utilize instant messaging (IM) applications such as Slack, WhatsApp, and WeChat as a communication channel among project participants to share daily construction information due to easy accessibility. However, general contractors are still required to manually extract and integrate the data from instant messages to compose daily reports. This is because the data inputted by subcontractors through IM applications are usually in an unstructured form and the IM application is not normally interoperable with the systems database especially developed for construction management. To solve this problem, this study proposes a chatbot-assisted construction daily report data management system. The chatbot in the proposed system collects and processes the required information through conversations with subcontractors, and automatically generates and shares a daily report for general contractors. A prototype system has been designed and implemented to prove the concept.

Prototyping a Chatbot for Site Managers Using Building Information Modeling (BIM) and Natural Language Understanding (NLU) Techniques

“…With the limited capabilities of self-implemented NLP, however, this study only supports simple sentences that contain no verbs or complex calculations, which limits the applicability of the proposed method. Cho et al [ 6 ] proposed a chatbot-assisted construction daily report data management system, which collects conversations among subcontractors on instant messaging applications and then automatically generates daily reports for general contractors. This study adopted the Natural Language Understanding (NLU) capability provided by IBM’s Watson Assistant to establish an NLU model that can understand the intents behind users’ messages for subsequent automatic extraction of the necessary information.…”

“…Another research [ 5 ] that aims to improve the efficiency and user experience of facility management also developed a chatbot that integrates NLP, Building Information Modeling (BIM), and ontological techniques to increase the efficiency of information retrieval. Cho et al [ 6 ] proposed a chatbot-assisted construction daily report system, which collects conversations among subcontractors on instant messaging applications and automatically generates daily reports for general contractors. This study adopted Natural Language Understanding (NLU), which can understand the users’ intents for subsequent automatic extraction of the necessary information.…”

“…Some studies [ 4 , 5 ] have leveraged instant messaging platforms to enhance the ease of use and usability of information systems, but they lacked Natural Language Processing (NLP), which resulted in poor performance. Other studies have adopted self-implemented NLP techniques [ 5 , 6 ] to extract information from textual data sources, but their lack of natural language understanding models results in high demand for manual input to maintain the corpus and limits their systems’ ability and applicability. While a few chatbots have been developed taking advantage of NLU models [ 6 ], those that also retrieve geometric data of buildings that are essential for construction site workers are still rare.…”

“…Other studies have adopted self-implemented NLP techniques [ 5 , 6 ] to extract information from textual data sources, but their lack of natural language understanding models results in high demand for manual input to maintain the corpus and limits their systems’ ability and applicability. While a few chatbots have been developed taking advantage of NLU models [ 6 ], those that also retrieve geometric data of buildings that are essential for construction site workers are still rare.…”

Conversation-Based Information Delivery Method for Facility Management

“…For instance, Tsai et al [ 45 ] developed a disaster management decision support system developed for the Bureau of Mine in the form of a chatbot. Cho and Lee [ 46 ] proposed a daily construction report data management system by using chatbot. The chatbot is responsible for collecting and processing the required information through conversations with users and generates daily reports automatically.…”