Deep learning for energy-efficient running of battery-electric trucks

To decarbonise the UK’s transportation sector, we must electrify long-haul heavy goods vehicle journeys. This project will investigate state-of-the-art control methods, including deep learning/machine learning, to determine the best way to control the world’s future battery-electric heavy goods vehicles to minimise the energy consumption of these vehicles and maximise their usable range. You will work with a team of experienced academic to produce research that is directly relevant to the decarbonisation of transport.

To reach the UK’s net zero pledge by 2050, the transportation sector needs to undergo a paradigm shift away from fossil fuels toward electrification. Even if trunk routes are electrified (e.g. high-throughput motorways), electric trucks will have to be able to undertake short trips away from electrification, during overtaking manoeuvres and when travelling between trunk routes and distribution centres or depots.

The energy required for these short hops away from electrification will be stored in batteries. This project will use advanced control methods, including deep learning, to determine the best way to use the energy in these batteries to ensure high energy efficiency while also minimising the wear on the batteries – increasing their lifetime. This will ensure that battery electric trucks can be used for a wider range of tasks, and require their batteries to be replaced less often.