SimplifAI

Lead Research Organisation: University of Huddersfield
Department Name: Sch of Computing and Engineering

Abstract

This project seeks to investigate the feasibility of the integration of a range of data sets about an urban area to enable transport operators to manage traffic flows more effectively, with particular application to the control of transport-determined air quality levels. The data sets include time varying data about traffic (average speed, flow rate), weather (wind speed, direction, temperature) and air quality (NOx concentrations), as well as static data about route topology, geography and infrastructural assets.

These data sets will be enriched to enable the use of an automated approach to derive regional strategies for urban transport operators, so that traffic flows can be influenced strategically, and in real time, through an urban region. Currently the creation of regional strategies by traffic operators is a time consuming manual process requiring a high level of experience and expertise, and is aimed specifically at minimising delay for traffic during exceptional events (such as road closures or large concerts).

The project will study the feasibility of a novel, two stage approach:

(i) exploring a new way of enhancing and enriching transport and environmental data feeds by adding semantics in the form of meta data and ontological context, to improve the clarity and usability of the data;

(ii) enabled by (i), provide automated support for the real-time creation of regional transport plans (strategies) that urban transport operators can enact.

The success of this study, and consequent exploitation of the technological advances, would lead to UTMCs being empowered to address the urban challenge of ensuring satisfactory air quality within urban areas, and will be able to inform specific road user groups (e.g. cyclists) of air quality conditions on their route choices. More generally, UTMCs will have access to technology to support the creation of regional strategies to address other challenges, such as dealing with road closures, multiple resource effecting scenarios, or balancing flows on the network, within real-time.

Planned Impact

Using a wide range of data about current and impending traffic flows, weather, the built environment and air quality readings, it is possible to predict an air quality problem within an urban region during a particular time period. What is not currently possible is to generate, in real time, specific strategies specifying changes in combinations of UTC assets (traffic lights, information signs, variable speed limits etc) that if carried out immediately are likely to avoid the air quality problem. This is the problem that we are attempting to solve, and if successful the feasibility phase has the potential to lead to widespread impacts. As well as the benefit on efficiency of strategy production, automation can lead to reduced errors in the process, and less reliance on in-house expertise.

Transport for Greater Manchester (TfGM) will benefit as they will gain a greater understanding of the sources and content of available data assets, and how they can enhance their service using this data. They will also increase their expertise in the construction of strategies for regional control of road traffic, and at the end of the project be able to utilise any tools delivered in helping to create the strategies for influencing flows of traffic through their region. In particular, they will be able to explore a means of influencing regional traffic flows in order to avoid compromising air quality standards. Partners BT, InfoHub and KAMfutures will have a greater understanding of the added value of heterogeneous data sets, their integration and enrichment, and the potential of centralised control strategies, to help in complex system management. Such is the ubiquitous nature of the problem of influencing regional flows of urban traffic, the products in terms of software tools and data sets which come about from the research have the potential to be sold commercially on a global basis.

UTMCs generally will benefit as they will have an examples of tools and methods that, taking advantage of a range of data, can be used in regional control. Starting with TfGM, the fusing of the data sets and development of the proposed technology will empower UTMCs to expand the service they provide to the community, and where that is used for air quality management, improve the quality of life for anyone using the Urban area. This would particularly help UTMCs where operator expertise is sparse. Drivers and the general public will eventually benefit as the use of the data and consequent more informed regional controls will better inform and control vehicles, better balance the network resource, and help to avoid pollution hot spots. Hence the research has the potential to improve the nation's health, by avoiding pollution build-up, and the nation's wealth, by commercialisation of any tools produced and data sets used to be sold in UTMCs worldwide. The timescale for impacts to be made will be approximately 3 years after the end of the feasibility, as the project will need to move through the demonstrator and full prototype stages before operational use.

Publications

10 25 50
 
Description During the project we demonstrated the integration of data feeds and their enrichment into logical facts using known data models and ontologies. These logical facts were used by automated planning technology to show how regional strategies (containing traffic light signals changes) in Urban Traffic Control in close to real time can be generated. The strategies were generated in less than 1 minute. We performed some tests to show how the strategies helped to lower tail-pipe emissions from vehicles.
We performed an evaluation on the strategies generated by visual inspection by urban traffic control experts, by internal simulation by planner execution and and by enacting the generated strategies using a micro-simulation models SUMO and AIMSUN. We compared the results against fixed strategies and showed that the planner-generated strategies led to a quicker clearance of vehicles from congested routes.
Exploitation Route The findings may be exploited by Road Traffic Operators. We are currently looking for funding to develop and use this technology to help lower harmful vehicle emissions in urban areas.
Sectors Digital/Communication/Information Technologies (including Software),Environment,Transport

 
Description The technical deliverables of this award are being evaluated for use in helping strategy generation for Urban Traffic Control Management, with potential contribution to alleviating traffic delay and minimising traffic related airborne pollution.
First Year Of Impact 2015
Sector Transport
Impact Types Policy & public services

 
Description First of a Kind (FOAK)
Amount £34,872 (GBP)
Organisation Innovate UK 
Sector Public
Country United Kingdom
Start 11/2016 
End 02/2017
 
Title Automated Strategy Generating Program 
Description We have created an intelligent programme which can input (a) a state of the road network in some urban region (b) a theory of traffic flow (c) a set of goals to succeed; and outputs an operational strategy detailing timings of traffic signals in order to achieve the goals. The goals can be linked to pollution levels within an Urban area, hence the generated strategy can help to lower pollution levels. 
Type Of Material Improvements to research infrastructure 
Provided To Others? No  
Impact This tool has been trialled in simulations of the Manchester urban area, and is the prototype for a system which we plan to trial with real tests in the Manchester area in the near future. 
 
Description SimplyfAI-2 
Organisation BT Group
Country United Kingdom 
Sector Private 
PI Contribution We were the Technical Lead - we supply the technical know-how and solutions.
Collaborator Contribution KAM Futures - project manager BT - technology suppliers Transport for Greater Manchester - problem owners InforHub - Traffic Simulation
Impact Submission and funding of First of a Kind stage 1 proposal Submission of First of a Kind stage 2 proposal
Start Year 2016
 
Description SimplyfAI-2 
Organisation KAM Futures Ltd
Country United Kingdom 
Sector Private 
PI Contribution We were the Technical Lead - we supply the technical know-how and solutions.
Collaborator Contribution KAM Futures - project manager BT - technology suppliers Transport for Greater Manchester - problem owners InforHub - Traffic Simulation
Impact Submission and funding of First of a Kind stage 1 proposal Submission of First of a Kind stage 2 proposal
Start Year 2016
 
Description SimplyfAI-2 
Organisation Transport for Greater Manchester
Country United Kingdom 
Sector Public 
PI Contribution We were the Technical Lead - we supply the technical know-how and solutions.
Collaborator Contribution KAM Futures - project manager BT - technology suppliers Transport for Greater Manchester - problem owners InforHub - Traffic Simulation
Impact Submission and funding of First of a Kind stage 1 proposal Submission of First of a Kind stage 2 proposal
Start Year 2016