Human-computer interaction specialists from the University of Glasgow are highlighting the need for new systems in autonomous vehicles (AVs) capable of replicating the complex social interactions between human car drivers and cyclists on UK roads.
Human-computer interaction specialists from the University of Glasgow are highlighting the need for new systems in autonomous vehicles (AVs) capable of replicating the complex social interactions between human car drivers and cyclists on UK roads.
It’s worthwhile research, sure, but as @sexy_peach@feddit.de said it won’t help on the short term.
Meanwhile, as long as there’s significant uncertainty on the vehicle’s ability to understand cyclists’ intent, there should be mitigations, like keeping a larger distance from cyclists, or protected bicycle lanes.
Also I would hope self-driving cars of a particular type won’t be allowed until they demonstrate robust abilities to detect and avoid collisions with pedestrians and bicyclists.
The priority rule of the planning software should always be to avoid a collision. Even if the car doesn’t read cyclists intent properly it should always be able to emergency brake. But to do this, it needs to be able to correctly detect and plan out other road users.