A self-driving motorcycle will allow autonomous cars to be tested under more challenging and representative conditions.
Motorcycles present a particular challenge to an autonomous vehicle or advanced driver assistance systems (ADAS) due to their speed and manoeuvrability.
To avoid risk to a human rider, initial development of the interactions between motorcycles and autonomous or ADAS-equipped vehicles has been carried out using controlled soft targets, but these are limited by their lack of speed and how accurately they represent real motorbikes.
Co-developed by Wiltshire-based AB Dynamics, the riderless motorcycle is said to have the full performance potential of the original bike, enabling more dynamic interactions to be tested, such as motorcycle overtaking, traffic filtering and lane splitting.
“A riderless motorcycle allows more comprehensive testing of autonomous or ADAS-equipped vehicles, without risking injury to a real rider,” explained Dr Richard Simpson, Senior Systems Engineer, AB Dynamics. “It also permits greater accuracy, repeatability and consistency between tests than any human rider could achieve.”
Mechanical integration of the riderless systems was carried out by technology start-up AutoRD. AB Dynamics said it integrated its on-board robot controller, which runs the company’s standard RC software, allowing programming of the motion of the motorcycle and path-following via GPS positioning. The bike, which can be coordinated with other moving objects such as cars or ADAS targets, synchronises all data generated for review.
The technology demonstrator used a BMW C1 which has ABS, no manual clutch and a roof structure where sensors are mounted. According to AB Dynamics, subsequent developments will use a more modern machine with greater performance.
“Future legislation and vehicle safety testing could require ADAS systems and autonomous vehicles to be validated in increasingly complex scenarios and the riderless motorcycle is a useful tool for achieving this,” said Simpson. “It could also have applications in motorcycle durability testing by removing the human rider from some of the more arduous tests over rough surfaces, such as pave, where cars already use robot drivers to eliminate driver fatigue.”
AB Dynamics said it looking for potential customers to help develop the project.