Volvo Cars is ushering its safety development program into the AI age. The company is equipping its newer cars with advanced sensors to collect more data than ever on emergency braking, sharp steering, manual intervention, and more.
The Swedish car company seeks to leverage AI to help identify where safety improvements can be made by collecting data from these sensors after accidents. This works similarly to aeroplane black boxes, which collect large amounts of data to help researchers identify what went wrong.
Volvo makes AI a safe(ty) bet
Volvo has long prided itself on its continuous research into vehicle safety. Technologies like ‘Side Impact Protection System’ (SIPS), and the ‘Whiplash Protection System’ (WHIPS), are a few of the car company’s safety innovations.
The company uses a 3D modelling technique called Gaussian splatting which uses data from the sensors to construct a 3D point map that can then be fed into 3D modelling software to recreate a highly accurate virtual model of the real-world scene. This will allow researchers to not only visualise accidents to see what failed but also to run prototypes through more accurate virtual stress tests.
In the age of big data, AI has been a godsend to developers for its ability to sort through large amounts, like looking for needles in a haystack with a metal detector. Back in the day, Volvo engineers would have to pull up to the scenes of car accidents, measuring the lengths of skid marks and the like. If the Swedish car maker was capable of producing innovations like SIPS and WHIPS with such crude data and tools, then this bodes well for its future safety endeavours.
