With regards to "can this ever actually be done?" I think it can*, but also it's got all the problems of both a software project and fundamental research: the difficulty is unknown, the timeline is a wild guess.
That said, no steering wheel? Finally, he's met my long-standing requirement for what counts as "genuinely self driving".
* humans don't need lidar — we clearly benefit from all the extra sensors or we wouldn't even have rear view mirrors let alone parking sensors, but technically it's believable.
That said, no steering wheel? Finally, he's met my long-standing requirement for what counts as "genuinely self driving".
* humans don't need lidar — we clearly benefit from all the extra sensors or we wouldn't even have rear view mirrors let alone parking sensors, but technically it's believable.