Ask HN: Whats happening with middle-mile autonomous trucks? | Hacker News Reader

Ask HN: Whats happening with middle-mile autonomous trucks?

3 pointsalgo_trader10 comments

This seems like an easy win. A lower technical barrier. Cost advantages (especially with e-trucks and slower/automatic charging).

Why is no company taking this market? Certainly no one with a tesla/nikola buzz.

View on HN

Yes, It does look more doable. I see many small truck pads where double trailers are driven off the interstate and parked - then a local driver(or two) come and decouple the pair and then locally deliver on the smaller roads. The driver who dropped them then picks up a pair of trailers to head down the line - or back? The mechanics of automating the hauler to drop and then pick up another pair seem simple. splitting the pair, letting down the legs, and building a new pair must now be done manually, air/power connections are done manually. The robo-truck needs to stop, set the air brakes, let down the legs automatically(now done manually) and pull forward and maneuver to pickup up and connect the next pair - with auto legs-up and air power connect and push into the third wheel connector. I suspect this is being perfected as we speak, it is so logical?

Actually, a pick-up/drop-off location will probably have a human present! So there is not even a need to fully automated this

How is the Tesla Semi (https://en.wikipedia.org/wiki/Tesla_Semi) not what you're describing?

Yeah there's plenty trying. Tesla Semi is going for L2, TuSimple and Waymo Via among others are trying L4.

But Tesla and Waymo are trying to solve the full problem.

If u limit yourself to an easier sub-set of pre-mapped routes, highways, speeds, weather, etc - maybe get a conditional L4 certification ahead of the others ??

And i think the business model still holds under these limitations. Just a thought....

How is it a lower technical barrier?

Pre-determined routes. Highway only. Start with with lower speeds. Easier to build charging stations/solar/etc. Lower engine performance is tolerable. Suspend driving during bad weather. Many more...

This is still L4 driving. Given that, why would you limit your approach to a smaller part of the market, given the same technical and regulatory challenges?

If u limit yourself to an easier sub-set of pre-mapped routes, highways, speeds, weather, etc - maybe get a conditional L4 certification ahead of the others ??

(Luckily, i am not a regulatory expert)

Also - trucks are bigger and more expensive, so more expensive and larger FSD hardware is possible??

Fair point - higher overall cost warrants higher sensor BOM costs. But I don't believe there is regulatory carve-out, so even if routes are set and pre-mapped (weather concerns are already built into L4) you still have the bulk of the autonomy problem to deal with.

Reply on news.ycombinator.com