Driverless trucks can now travel farther distances faster than human drivers | Hacker News Reader

Driverless trucks can now travel farther distances faster than human drivers

techcrunch.com

25 pointsjimt123417 comments

Due to federal safety regulations. I wonder how much safer this is?

> Revenue is projected to continue as the company adds more trucks and driverless routes to its network. Today, the company has 30 trucks in the fleet, 10 of which are operating driverlessly. That fleet is expected to grow to more than 200 trucks by the end of the year. Urmson said the company’s trucks have racked up 250,000 driverless miles as of January 2026 with a perfect safety record.

> In the second quarter, Aurora plans to deploy a fleet of driverless International Motors LT trucks, which will not have a human observer on board. Aurora’s driverless operations that use Paccar trucks currently have a human safety observer in the cab as requested by the truck manufacturer.

I don't know that there's enough data here to say that their current safety record justifies the claims just yet. If it's not safer than overworking a regular driver then the headline claim is just regulation arbitrage.

250,000 miles is a small number when talking about average drivers, let alone professional. My mother has driven atleast 200,000 miles in her life and never been on an accident. Ive done 100,000 miles on a motorcycle with only two hair raising close calls, and bikes are 60x more dangerous than cars mile for mile.

I don't think you can really get any significant data from 250,000 miles because most people don't wreck in 250,000 miles, especially professionals.

This checks out. "Miles Between Accidents: Based on an average rate of ~0.74 per million miles, this equates to roughly one accident for every 1.35 million miles driven."

And how many hundred thousand truck drivers are now facing the end of their careers?

Right now, none.

Aurora seems to be flying under the radar, this is the first I've heard of them. They seem to be actually deploying trucks, and not just smoke and mirrors.

I have long thought that fixed distance trucking is the most approachable self driving outcome. Do not attempt the billion messy real world scenarios which occur in urban environments when surrounded by humans.

Make a few hubs that are well outside of city centers that are connected by nothing but straight interstate. AI can drive the truck on this boring fixed route. Humans then take over the short distance to bring it to the actual delivery location.

congratulations! you're invented an inefficient railway

The US already has the best freight rail network in the world my many measures but you can't put every shipment on a train.

Do you think long haul trucking does not exist today? Use a train when you can, but that is not always an option.

There's no operating rail between LA and Phoenix, freight or passenger. Union Pacific rerouted freight away from Phoenix decades ago, Amtrak followed suit, and the previously used rail line that connected LA and Phoenix is out of service. See https://en.wikipedia.org/wiki/Phoenix_Subdivision_(Union_Pac... and https://www.arcgis.com/apps/mapviewer/index.html

The highest volume OTR shipping lane is LA to Phoenix, which is already the perfect place for self driving vehicles.

I've been saying for years, trucks should drive autonomously from one mega parking lot outside a city to another at nighttime, and have humans handle the last mile during the 7-3 shift.

Do these trucks actually drive themselves into a warehouse, and back into an empty spot? Can they actually navigate cities? Or only on the interstates? What kind of loads are they carrying? Can they navigate mountain passes? How much of the entire trip is actually automated? They never answer any of these questions, and make it sound like they can drive anywhere, any time. But it seems to me more like they are on predefined routes, limited to certain destinations. Maybe even someone picks up the truck at an off ramp and drives it into the city. That's not a bad thing, but they are so sketchy the way they don't give any details. It makes it hard for me to believe half this stuff.

> Can they actually navigate cities? Or only on the interstates? What kind of loads are they carrying? Can they navigate mountain passes? How much of the entire trip is actually automated? They never answer any of these questions, and make it sound like they can drive anywhere, any time.

See Waymo. Answering these questions will make the investors run away.

Basically, a highway is the best place for selfdriving cars. Cities are difficult even for selfdriving cars (Waymo).

Imagine Tesla showed off a semi almost 10 years ago and they have some capacity of self driving. They could've done this years ago if it wasn't a made up timeline. Good work to these people. I don't know how they're spending that much money per year, but the hardest part is starting and they've started. High hopes

It’s a bit unclear what they’re actually doing. Since they need to have safety drivers for now, it sounds like they can’t actually do this nonstop yet and they have only “validated the route?” Or are regulations more relaxed for safety drivers?

AUR stock down 36% over the past 6 months; down 52% over last 12 months.

Reply on news.ycombinator.com