AbstractIntelligent Transportation Systems (ITS) have been considered as a solution to the problems associated with modern urban traffic management systems. The implementation of ITS is highly complex due to the multi-domain and geographically distributed nature of the traffic controls/systems and legacy infrastructures. Some of the approaches that have attempted to address these challenges include control engineering, artificial intelligence, telematics, grid computing and multi-agent systems. This research work was undertaken to investigate the implementation approaches aimed at addressing the complexity and communication challenges of ITS, specifically intelligent and cooperative traffic systems/controls which are capable of distributed decision making in response to emergent traffic situations.
This research introduces the novel concept of Semantic Agent-based Controls which utilise a unique combination of computing concepts including multi-agent systems, Service Oriented Architecture (SOA) and semantic web services. The semantic agent-based controls extend the functionality and decision making capabilities of fixed traffic controls by using semantic agents over the network to effectively control geographically distributed traffic controls/roadside systems.
The research introduces a novel Agent communication and coordination mechanism that facilitates the interaction between the Semantic Agent-based Controls in a multi-domain ITS environment. The mechanism uses a semantic layer approach based on the principles of SOA and the concept of semantic web services which enables Agents to store, share and integrate heterogeneous data on the semantic level across the SOA enabled network of traffic systems and devices. This highly flexible mechanism allows semantic agent-based controls to interact with each other by using different levels of ontologies that semantically describe the Agents’ intentions, behaviours and coordination strategies. The semantic communication layer allows coordination of action plans in the form of decision rules between the semantic agent-based controls which facilitates dynamic and distributed decision making in response to emergent situations.
A new ITS platform (named ITS@CU) is presented, which was developed to support and evaluate the semantic agent-based controls approach. The ITS@CU platform (and the associated simulation utilities) was developed using the latest state of the art commercially available tools and technologies. It is a multi-agent platform and is based on advanced SOA principles which enable seamless integration of traffic controls/systems.
The research approaches are evaluated using a variety of different test cases generated using historical traffic data and typical traffic situations identified by UTMC Control Room, Coventry City Council. The results demonstrate the advantages of Semantic Agent-based Controls especially when compared with traditional controls (without Agents). The advantages include communication reductions, flexible and robust coordination between the distributed controls, and enhanced decision making and self-organisation capabilities.
The research presented started as a Knowledge Transfer Programme (KTP) funded project in collaboration with T@lecom Limited. Elements of the ITS@CU platform developed during the project also led to the design and development of the following three successful commercial applications/products for T@lecom. The “Patient Transport System” is a mobile solution for ambulance crews to transfer patients to and from hospitals and with novel features of intelligent tracking, two-way messaging/alerts and patient prioritising in real-time. The “Mobile Gateway System” is a SOA enabled unified communication and integration platform for T@lecom’s mobile solutions. Finally, the “Vehicle Tracking Solution” incorporated comprehensive functionalities of the ITS@CU platform and also includes a novel driver behaviour assessment.
|Date of Award||2011|
|Supervisor||Olivier Haas (Supervisor)|
- Agent-based Controls
- Urban Traffic Management and Semantic Web Services
- Grid Computing
- Multi-Agent Systems