Codifying information flow in ‘construction as a manufacturing process’ in BIM-enabled projects to support future automation.
Ranjith K Soman - PhD candidate 2017 - present
This PhD study aims to automate the generation of Look-Ahead-Schedules (LAS) for modular construction projects, making use of opportunities offered by state-of-the-art developments in machine learning and tracking technologies. Vast amount of data, complex and dynamic constraint relationships, multiple work fronts etc. makes the manual generation of LAS slow and inaccurate. Automating the generation of LAS needs information associated with the construction processes, which are fragmented across domains, to be codified and integrated. This requires a new method for modelling process information. To develop the information modelling approach, codification challenges and practices in the execution of projects must be studied in detail. In the current PhD, an effort is made to 1) understand the construction codification challenges through an interpretive case study, 2) a new information modelling approach for codifying process information is developed, building on the researches on linked data in the built environment and 3) the LAS generation is automated using reinforcement learning and dynamic constraint solving. The PhD is supervised by Prof Jennifer Whyte and co-supervised by Dr Miguel Molina-Solana (from 2019) and Dr David Birch (February 2017-December 2018). Industrial supervision for this PhD was carried out by Dean Bowman, Bentley Fellow, Bentley Systems.
Aim and Objectives
The aim of this research is to develop an information modelling method to codify construction process information in BIM-enabled projects to support look ahead planning automation
- To understand the codification challenges in the late-design and construction phase of BIM enabled projects;
- To develop an information modelling approach to codify construction process including dynamic constraint information to support future automation; and
- To develop a method to generate look-ahead-schedule for modular construction projects automatically using the codified process information.
The research is being done at Centre for Systems Engineering and Innovation and is funded through Skempton scholarship and co-sponsored by Bentley Systems. Ranjith spent 6 months (October 2018-March 2018) at the The Alan Turing Institute using the PhD enrichment scheme at Turing.
• Soman, R.K. (2019) Linked-data based dynamic constraint solving framework to support look-ahead-planning in construction, 36th CIB W78 2019 Conference on ICT in Design, Construction and Management in Architecture, Engineering, Construction and Operations (AECO), 18–20 September, Newcastle, United Kingdom.
• Soman, R.K. (2019). Modelling construction scheduling constraints using Shapes Constraint Language (SHACL), European Conference on Computing in Construction (EC³ 2019). Crete, Greece.
• Whyte, J.K., Mahalingam, A., Soman R.K. (2019). Temporality, innovation and megaproject-to-megaproject learning across continents in the case of Crossrail and Nagpur Metro. 35th European Group for Organizational Studies (EGOS) Colloquium. Edinburgh, UK .
• Senthilvel, M., Soman, R.K., Mahalingam, A., Whyte, J.K., Raphael, B., Brilakis, I., & Varghese, K. (2018). Towards digital delivery of metro-rail projects in India. 7th World Construction Symposium 2018: Built Asset Sustainability: Rethinking Design, Construction and Operations. Colombo, Sri Lanka.
• Soman, R.K., Birch, D., & Whyte, J.K. (2017). Framework for shared visualization and real-time information flow to the construction site. 24th European Group for Intelligent Computing in Engineering (EG-ICE) International Workshop on Intelligent Computing in Engineering. Nottingham, UK.
• Soman, R.K., & Whyte, J.K. (2017). A framework for cloud-based virtual and augmented reality using real-time information for construction progress monitoring. Lean & Computing in Construction Congress (LC3). Heraklion, Greece.