Real-time Intelligent Map-matching Algorithms for Advanced Transport Telematics Systems (RiMATTS)

A variety of transport applications and services such as pay-as-you-drive insurance scheme, navigation and route guidance, accident and emergency responses (enhanced 999 emergency services), bus arrival information at bus stops and fleet management require spatial and temporal location, and time information. One of the important components of such services is the navigation module which provides the required positioning data. Many commercial devices are available to support navigation modules of such transport systems. In recent years, most commercial devices use GPS technology for acquiring such positioning data. Since GPS suffers both systematic errors and noise, the required positioning accuracies of many transport services cannot be achieved by such devices. Moreover, such devices do not provide integrity (the level of confidence) of position solutions which is very important for liability and safety critical applications such as pay-as-you-drive insurance schemes (due to the possibility of billing incorrectly) and responses to emergency 999 calls. A map matching algorithm that integrates the locational data (from GPS or other sensors) with the spatial road network data needs to be employed. Map matching not only enables the physical location of the vehicle to be identified but also improves the positioning accuracy if a good digital map is available. Current map matching algorithms are not capable of supporting the navigation modules of certain transport systems in some operational environments (specifically in dense urban areas) due to the inherent limitations and constraints associated with them. In addition to this, a single map matching algorithm cannot optimally support the navigation module of a transport system in different operational environments. Therefore, there is a distinct need to select a set of representative map matching algorithms. The detailed characterisation of these algorithms through experiments is essential to evaluate their performance in the operational environments for which they were designed and to identify their limitations. This representative set of existing map matching algorithms with further enhancements, along with a new map matching algorithm that can take into account limitations and constraints of existing map matching algorithms, could optimally support the navigation modules of most transport systems in most operational environments. Therefore, the main objectives of this research project are to (1) identify a set of representative map matching algorithms from existing algorithms, (2) develop a new map matching algorithm and to address any gaps identified in objective 1 both in terms of applications and operational environments, (3) develop a knowledge-based intelligent map matching (iMM) technique to identify the best map matching algorithm (achieved in objectives 1 and 2 above) suitable for an operational environment, and (4) demonstrate a potential application of iMM in different operational environments. Several criteria will be defined for use with the iMM technique to select the best algorithm for a particular service in a given operational environment. Such criteria will include the geographic characteristics of the operational environment (such as land-use, road network density, and building height information) and others (such as complexity, and cost) if required. The exploitation of this proposed research would be in two levels: (1) the algorithms, (2) the actual navigation system which incorporates the algorithms and the navigation sensors. The expectation is that the cost associated with the actual navigation system will be relatively low (at the level of 500 per unit). This is expected to fall as the price of navigation sensor chips and MEMS technology-based sensors reduce over time.