Cloud-based traffic data fusion for situation evaluation of handover scenarios
This addresses safety challenges in conditional automation for drivers who may be inattentive, though it appears incremental as it builds on existing data fusion approaches.
The paper tackles the problem of safely transferring control from an automated vehicle to a distracted driver in urban scenarios by fusing cloud-based traffic data from infrastructure, vehicles, and drivers to evaluate handover situations, aiming to enable timely and possibly pre-critical transfers.
Upcoming vehicles introduce functions at the level of conditional automation where a driver no longer must supervise the system but must be able to take over the driving function when the system request it. This leads to the situation that the driver does not concentrate on the road but is reading mails for example. In this case, the driver is not able to take over the driving function immediately because she must first orient herself in the current traffic situation. In an urban scenario a situation that an automated vehicle is not able to steer further can arise quickly. To find suitable handover situations, data from traffic infrastructure systems, vehicles, and drivers is fused in a cloud-based situation to provide the hole traffic environment as base for the decision when the driving function should be transferred best and possibly even before a critical situation arises