Naghshbandi, Seyedeh Neda; (2022) Technology Capabilities for Safe and Resilient Coordination of Automated Earthwork Systems: A Decentralized Multi-Agent System Approach. Doctoral thesis (Ph.D), UCL (University College London).

Text ThesisUCLJuly2022-Final.pdf - Other Access restricted to UCL open access staff until 1 August 2025. Download (5MB)

Abstract

A systematic examination and evaluation of operational system safety is required for automated earthwork systems due to its distinguishing characteristics compared to other automated industrial systems like the manufacturing and mining industries. Earthwork systems operate in an unpredictable and uncontrolled environment. Accordingly, complete safety cannot be assured, yet we can progress toward it using digitisation and object detection technologies, and efficient coordination strategies. As such, the problem is: How do we enhance the future system safety of complex automated earthmoving operational systems that a digitised construction industry relies on? This thesis centres on operational systems’ safety, and consequently the impact of technology innovations on emerging safety risks from the interaction between automated earthmoving equipment units and operational environments were studied. A systematic literature review was conducted to identify the existing technology innovations and their capabilities in enhancing earthworks’ system safety. Then a hybrid distributed multi-agent system was developed to analyse an earthwork’s system safety continuously via simulating the interaction of automated earthmoving equipment units with the environment during their operation. The developed hybrid model integrates Agent-Based Modelling and Discrete Event Simulation methods. The model developed for this thesis evaluates an earthwork’s system safety through assessing both its safety, and productivity concurrently. The developed hybrid model contributes to better safety management and planning of complex earthwork systems in a digitized construction industry, as well as further development of earthwork operation agent-based modelling paradigm as it is a tool to support safe behaviour generation for agents such as automated earthmoving equipment through: 1- Evaluating and analysing both the system’s safety and productivity concurrently, based on the estimated state of earthmoving operations; and quantifying them. 2- Evaluating the efficiency of plans and actions. In fact, it facilitates defining better safety enhancement strategies by tracing the states and attributes of agents, situations, events, and interactions to acquire the required information and knowledge to control automated environment and associated agents’ actions and plans their future behaviour. 3- Testing the effect of individual agent’s actions (e.g. path planning) on fleet-level coordination of the earthwork system. In other words, this thesis contributed to the body of knowledge by establishing a distributed multi-agent framework for enhancing an earthwork system’s safety without compromising its productivity through applying agent-based modelling which has not been sufficiently utilised in this area of science. A key strength of the hybrid model is its scalability and flexibility. This can be used to accommodate a wide range of excavation site requirements, including different layouts and parameters. This study shows although historical data analysis can define safety barriers, this cannot be completely responsive to uncertainties related to a complex system. Highly automated systems (autonomous systems) require real-time reactive capability to deal with dynamic and uncertain interactions within the operational system. In other words, automated earthwork systems require a high degree of adaptability to ensure safety. Higher adaptability can be achieved through integrating 1- accurate object detection technologies, 2- appropriate coordination strategies, and 3- operational environment modifications. Even though this study is able to test a wide range of scenarios, it does not guarantee a safe, resilient performance under all possible circumstances that could arise from equipment-environment interactions. However, simulation of scenarios clarifies the consequences of the defined actions for different operational states and shows how automated earthwork systems are safe and productive.

Download activity - last month

Export as JSONCSVXML

Download activity - last 12 months

Export as JSONCSVXML

Downloads by country - last 12 months

Export as CSVXMLJSON

Archive Staff Only

View Item