Comment by losteric | Hacker News Reader

Comment by losteric | Hacker News Reader

Parent Full thread

Well even Germany isn't going to start with self-driving trucks driving everywhere. Every country has standardization problems... in Germany there are ancient towns with arches that trucks can't go under (and not all of them are marked). In the US we have inconsistent/missing signs, dense urban areas built for small cars, and all kinds of crazy human driving.

My guess is actually opposite of yours though. I think short distance trucking will come last, with long-haul interstate trucking coming first. Short distances don't really need an automated truck... the trucker just gets in, drives, and arrives. Automating/standardizing loading/unloading would be great but automating the driver is less of a win.

On the hand, take interstates. Trucker gets in and drives for upward a few days with multiple mandated stops for naps/food... afaik, it takes 3 days to safely drive from SF to NY because of those stops. Now imagine a self-driving truck where the driver is only responsible for first/last-mile driving and refueling... done in 48 hours flat. 30% shorter transit time, plus whatever fuel/insurance savings the AI driver generates. Not bad.

> ancient towns with arches that trucks can't go under (and not all of them are marked).

And that are routinely blocked by trucks with human drivers who keep forgetting their required height clearance. If there was a way to make shipping companies fully responsible for all the delays caused but stuck trucks, laser-scanner assistance systems that stop the truck before it is stopped by the arch would already be a popular retrofit.

Trucking as an industry is still firmly entrenched in the 20th century. Those shipping companies do have to pay for delays, but the industry's mindset is that delays are an immutable cost of doing business... get insurance against delays, travel around areas that might pose risks (even if it takes longer), train humans to predict/handle exceptional cases, pad out shipping schedules, etc. There's plenty of room for disruption.

Truck manufacturers are onboard because they get to sell replacements for every truck on the road... it's the shipping companies that are going to be the hard sell. The market is ripe for disruption but both shipping customers and shippers will be hesitant when faced with new fangled laser-thingamajiggers.

To be honest, the haulage company or their insurance company is losing money whilst their truck is stuck under a bridge, and even more when someone sends them the bill for the bridge repair. If laser scanners or databases attached to geolocation services aren't sufficiently accurate or affordable to make even that minimal level of driver aid widespread, it's probably an indication that adequate driver replacement services are going to be even further behind for those routes.

The trucks will likely drive at 45 mph so it would take 65 hours to cross the country. Still a huge improvement.

> The trucks will likely drive at 45 mph so it would take 65 hours to cross the country. Still a huge improvement.

By "likely" you mean one guy in a Tech Crunch editorial mentioned that would be efficient. Driverless trucks could look totally different because they don't need to house a driver (and their sleeping quarters) and be much more aerodynamic.

Reply on news.ycombinator.com