Personalised smart vehicle technology

Wen-Hua Chen, Cunjia Liu and Dewei Yi have developed an advanced driver assistance system (ADAS) that learns the behaviour of a specific driver in particular circumstances, making it safer and more responsive than current generic systems.

Their ADAS utilises a personalised learning framework (PLF) and clustering-aided new user approach.

The PLF learns an individual’s driving characteristics. Unsupervised learning algorithms auto-tag and learn the driving behaviours, and then supervised learning algorithms recognise the individual’s driving status.

To address the challenge of the lack of data about a new driver, clustering-aided new user approaches extract features from other drivers with similar backgrounds or experience.

The new system has been successfully implemented, tested and validated through real-world datasets, demonstrating the benefits of customised vehicles for individual drivers.

The proposed personalised smart vehicle technology is easy to implement and relies on only basic vehicle status sensors and measurements without the need for cameras to monitor behaviour.

The system has a wide range of applications, including supporting intelligent vehicles to recognise the driving styles of other road users to better understand and negotiate with them, and making intelligent vehicles more suitable for individual driver’s needs.

The research has been published in four journal papers and as several conference papers.