On the road to efficiency

By optimising loads, routes and distances travelled for trucks, challenges with traffic congestion, costs, timings and environmental pollution can be tackled. Han La Poutr?© at Centrum Wiskunde en Informatica (CWI) is developing software to intelligently organise the road haulage business.

The DEAL project (Distributed Engine for Advanced Logistics) has studied the trucking business in the Netherlands in order to make it a lot more efficient. Traditionally trucks do not fill up to capacity but often complete return trips or have to undertake trips with either very small loads or with no load at all. No doubt this inefficiency only adds to the Netherlands’ notorious traffic congestion problems, with a proliferation of trucks on the roads.

When booking a load many deals are done one by one in turn and organised by people trying to get an approximate overview of all the loads coming up or available and so opportunities to make the most of transporting combinations of loads for optimal effect are lost.

The DEAL project has created a system that relies on autonomous software agents to take over the processes of organising, and bidding for loads and planning the routes to and from for maximum efficiency. With GPS/GPRS, technology to track vehicle positions, RFID to trace packages and supply chains connected together using broadband connectivity and XML interfaces, a state of the art overview of the whole relevant transportation system can empower the intelligent technology and the people using the system with more complete information for quality decision making.

Han La Poutré of DEAL explains the concept: “Something like 30 per cent of the trucks on the roads are currently empty. For the partial truck loads it’s important that the truck is filled sufficiently and the question for the project for the CWI was – how can you fill loads from all these partial truck loads but also how can you anticipate future loads to fill these trucks up later on?

“When everything that will be available is known far ahead in advance before you start driving you can already calculate what would be a good combination.”
The software modules, which are under continuous development, can bid intelligently for a truck load and if the software ‘agent’ makes the right decision the truck can drive making the most profit. It can calculate shortest routes, price considerations, load capacity and do this automatically. It even learns from itself about the best decisions made historically using so-called ‘genetic algorithms’.

With less trucks on the road due to the increased efficiency, emissions will obviously be reduced, making this not only a sound commercial endeavour but also an environmentally friendly one.
“There is one other direction with the project – a framework we made for outsourcing transportation jobs,” adds Han La Poutré. “If you, as a big company, cannot do it yourself efficiently you can continue by outsourcing your jobs to smaller companies with an auction mechanism.”

Published: Friday, 6th November 2009

Comment on this article

Comments

No comments for this article, be the first to comment