Project: Assessing the Potential for Rationalising Road Freight Operations.
Reference: STP 14/6/11
Last update: 27/07/2004 13:48:23
The project will develop a software tool-kit that can be used to interrogate an existing database containing information on 7000 road freight deliveries made by roughly 2500 vehicles over a 48 hour period in October 1998.
The tool-kit will be used to assess the potential for reducing the economic and environmental costs of road freight transport by improving backloading, increasing load consolidation, delivery rescheduling and more direct routing. The project will also investigate ways of combining the software tool-kit with real-time road freight information systems to identify backloading and load consolidation opportunities on a short-term basis.
UNILINK Unit, Technology and Research Services, Edinburgh, EH14 4AS
+44 (0)131 451 3071
Cost to the Department: £46,920.00
Actual start date: 25 July 2002
Actual completion date: 05 March 2004
Assessing the Potential for Rationalising Road Freight Operations
Author: Professor Alan Mckinnon
Publication date: 30/11/2003
Source: LOgistics Research Centre, Heriot-Watt University, Edinburgh.
More information: http://sml.hw.ac.uk/logistics
Summary of results
- Background and Aims
A key element of the government's Sustainable Distribution strategy has been the Transport KPI initiative. The main aim of this initiative has been to help companies benchmark the efficiency of their road transport operations against a standard set of key performance indicators.. Several 'synchronised audits' of fleet activity have been undertaken since 1997, including three in the food supply chain. Earlier analyses of the databases generated by these surveys have been confined to inter-fleet benchmarking and failed to exploit the opportunity for transport modelling afforded by the post-code data available for journey origins and destinations. It was recognised that analysis of this spatial data could permit an assessment of the potential for rationalising the distribution of food in the UK under varying conditions. This main objective of this project has been to develop software tools to assist this analysis and to apply them to the transport KPI data collected in a survey of 53 vehicle fleets (approximately 3500 vehicles) in the food sector in May and October 2002. Subsidiary aims have been to examine the collection and analysis of KPI data by commercial vehicle telematic systems (VTS) and investigate ways of combining the software tool-kit with real-time VTS to identify backloading opportunities.
Two sets of software tools have been developed. The first set interfaces the transport KPI data-base with the commercial vehicle routing package, Optrak. Individual trips are reconstructed and the routes the vehicles actually followed compared with optimal routes that they could have followed under varying scheduling constraints. The second set assesses the potential for increasing the level of backloading. They extract relevant data from the Access database containing the transport KPI data, reformat it, check it for consistency, geocode it and finally undertake a load matching analysis within various constraints using the GIS modules in the SAS software package.
Analysis of Routing Efficiency
Across the fleets sampled, between 5 and 10% of delivery and collection points had missing or anomalous post-codes. As the delivery rounds analysed had an average often collection / delivery points, there was a high probability of a route having an inadequate set of post-code data. In an effort to maximise the sample of routes, considerable time was expended checking and, where possible correcting, post-codes. Within the available time it was possible to check and analyse the routing data for a total of seven fleets. Over the 48 hour survey period they made a total of 469 trips for which adequate post-code data were available.
A comparison was made between the actual distances travelled and transit times, as recorded in the transport KPI survey, and the optimum values that could be achieved under varying conditions. Six scenarios were constructed in which three key scheduling variables were altered: length of the drivers' shift, delivery time window at customers' premises, opening and closing times at these premises. The modelling revealed both the potential for reducing traffic, cost and emission and the sensitivity of these savings to scheduling constraints. As the KPI survey did not enquire about the flexibility of delivery schedules, it is not known to what extent current trading and operating practices would have to be changed to realise these savings.
Analysis of Backloading Opportunities
Recent trends in empty running were examined and previous research on backloading opportunities reviewed, principally that of Cundill and Hull (1979). The analysis of back loading potential using transport KPI survey data overcame some of the limitations of this earlier study. The first step in the backloading analysis involved finding empty journey legs that could potentially be allocated backloads. Various levels of screening were applied to these potential loads relating to (a) location and direction of freight movement (b) vehicle compatibility (c) vehicle capacity and (d) delivery schedules. An interactive query interface was created using the GIS modules in the SAS software package to allow the user to identify backloading opportunities on a company, zonal or individual trip basis. Across the 29 fleets selected for backloading analysis, 573 empty journey legs were identified for load matching. By matching load origins and destinations it would have been possible to eliminate approximately a third of these legs (181). Each of these matches would save an average of220 vehicle kills. Empty running would have been reduced by 13.7%. Screening for vehicle compatibility, vehicle capacity and scheduling reduced the number of load matches cumulatively by 48, 72 and 47. This meant that only 14 potential load matches met all four sets of backloading criteria, eliminating only 2.4% of all the empty legs longer than 100kms.
The transport KPI survey was designed to collect standardised vehicle operating data that could be used to benchmark the efficiency of companies' road freight deliveries. As routing and backloading analysis was not in the original specification of the survey design, it is hardly surprising that the content and structure of the resulting data-set is not ideally suited to this type of research. The project reviewed the limitations of the transport KPI data and explained how they might be overcome. The main limitations included the poor quality of much of the post-code data, failure to monitor the activities of the tractor unit and lack of data on delivery time windows. The final report also discusses the transferability of the software tools to data collected in transport KPI surveys in other industrial sectors.
Surveys of Suppliers and Users of Vehicle Telematics Systems (VTS)
Two samples of 33 suppliers and 32 users oflorry tracking services were surveyed by telephone interview to examine (i) the potential for vehicle telematics systems to collect operational data required for the calculation of transport KPls (ii) the nature, formatting, storage and analysis of data currently collected and (iii) the interfacing ofVTS databases with other software packages. At present, te1ematics services are designed mainly to help companies improve time utilisation of vehicle assets, fuel efficiency and driver productivity. The current inability of VTS to monitor vehicle loading is a major constraint on the wider of adoption of the vehicle utilisation measures promoted by the government's transport KPI initiative. As load data is a key element in the spatial analysis of the transport KPI data, the applicability of the software tools developed by this project to commercial road telematics data-bases is currently limited.
Departmental Assessment Status: Assessed by FIT Programme Advisory Group 15 July 2004. Score allocated 7/10 - Very good project, achieved targets and objectives, good quality research work, outputs with commercial potential.