Issue date: 13 March 2003

ISSUE DATE: 13/03/03

Aggregates such as sand, gravel and crushed rock form the mainstay of construction materials for buildings and roads. Assessing the suitability of materials for different tasks is determined using different test methods. One important method is called petrography which is a time consuming manual process carried out by highly trained geologists, in order to acquire information on the petrographic composition of the aggregate particles. Engineers from the University of the West of England are working on a collaborative project to devise an automated vision based method capable of carrying out this analysis more speedily but just as efficiently as manual petrography. The project will involve the determination of fundamental properties of the particles, namely size and shape.

A prestigious research contract, for £230,000, has been awarded by the Department of Trade and Industry through the Eureka initiative, the pan-European network for market-oriented, industrial R&D, for this international collaborative project entitled ‘Petroscope’. The aim of the project is to develop an automatic, optoelectronic testing instrument for process and quality control in the construction aggregates industry. Dr Melvyn Smith, Dr Sagar Midha and Dr Lyndon Smith from the Faculty of Computing, Engineering and Mathematics at UWE, will use the funding to devise a part of a mechanised instrument that will analyse construction aggregates to manual petrographic test standards. The analysis of construction aggregates using machine vision is academically challenging and has enough scope for research at PhD level. The research associate working on the project will therefore be offered the opportunity to register for a PhD.

The UWE Engineers will do this using an automated vision based method which involves using specialised lighting, a camera and a computer. An aggregate sample is presented in front of a camera so that size, three dimensional shape and surface texture can be measured.

Currently no automated instrumentation exists that is able to analyse sand, gravel, crushed rock and other construction aggregates anywhere near to classical manual petrographic standards. The development of new forms of vision-based techniques offers significant opportunity for improvements in the operational performance of process and quality control for construction materials and a reduction in the environmental impact of quarrying.

“It is important for construction contractors to be able to estimate how different foundation aggregates such as sand, gravel and crushed rock will compact so that the right materials are used for the job in hand,” said Dr Melvyn Smith. “Improvements in materials for construction follow from an ability to better describe the properties of component mineral materials. Shape and size distribution and petrographic composition has a huge influence on the way materials pack together, the strength of concrete made from mixtures of different aggregates or the wear properties of a road.”

“This project builds on our previous innovative research work in vision-based three-dimensional data capture and surface topographic materials. We aim to work towards devising an automated vision based method both for the characterisation of three-dimensional size and shape distribution and the analysis of surface texture for complex particles with random pose. “

Dr Sagar Midha said, “The technology developed through this project could have far reaching benefits in terms of reduced cement and water usage for building construction and reduced bitumen usage for road and pavement construction. The environmental impact of these savings is highly significant.”

Dr Lyndon Smith said, "The UWE contribution to the Petroscope project represents an exciting opportunity for research into machine vision techniques for analysis of stone particle morphologies. This will provide the aggregates supply industry with a new technology that will enable the realisation of significant economic benefits.”

The researchers will be working with international partners including Petromodel Ltd. (Iceland; the main partner and initiator of the project), RMC Aggregates (UK) Ltd, Steinbeis Transfer Centre for Data Analysis and Visualisation (Austria) and ZAG - Slovenian National Building and Civil Engineering Institute.


Back to top