Just two weeks after Uber announced that it is closing down its self-driving truck division, a new startup has revealed itself to the world. And they’ve already got $40 million in funding.
Kodiak Robotics is similar to Uber’s self-driving truck program in a few ways. Uber’s program began as a startup called Otto. One of Otto’s cofounders was a former Google employee named Don Burnette. Burnette is also one of Kodiak’s cofounders. Otto (and later Uber) was trying to create a kit which could retrofit an existing truck to give it self-driving capabilities. While details around Kodiak are still hazy, it looks like they’ll likely be attempting the same thing.
Kodiak was founded in April, but stayed in stealth-mode until this week when it announced that it had completed its $40 million series A funding round. The company will use the money to hire more staff and start its test fleet. Currently Kodiak only has around 10 employees.
Even though Kodiak’s origin story sounds a lot like Otto, Burnette says the new company is different.
For one, Kodiak will likely be using LiDAR sensors, radar, cameras, and sonar – just about every type of sensor possible. All of that technology already exists. Now the challenge is to design a system that can take all of that information and know what to do with it.
Burnette also says that unlike Uber, Kodiak will be using sensors built by other companies – a plan which could keep them from ending up in the same legal trouble as Uber did.
And finally, Kodiak will be focusing 100% on autonomous trucks. No cars.
Burnette said. “A company has to be 100% hyper-focused on one problem and one problem only. We plan to solve just the long-haul, self-driving truck problem.”
As for how to solve that problem, Kodiak is likely planning to go the highway-only-driving route with a human driver taking over on local roads.
“You get on the road, stick in the right lane, and don’t hit the thing in front of you,” says Burnette.
Source: gobytrucknews, truckinginfo, ttnews, wired, techcrunch, forbes
Gil James says
Get on the road. Stick in the right lane and don’t hit the thing in front of you..Sounds easy. If it were that simple, no truck driver would ever have an accident. Hi tech people creating self driving trucks are just like transportation managers and dispatchers. Most have never driven a trucking unit and have no idea what it takes. The problem with technology is that it is so black and white. There is not enough grey area for skills, experience and most important, road awareness..
David Obermark says
I am a truck driver with my own authority who has over 20 years experience and over 2 million miles throughout the 48 CONUS states. I am now semi retired to just a part time contract where I do not even need a logbook let alone an ELD.
I agree with Mr Burnette that at least in the short term, the only potential for self driving trucks is for long haul trucking on freeways, tollways and perhaps other highways that although not limited access are not congested. I also agree a human needs to be in the driver’s seat ready to immediately take over when the need arises.
Yes, this will result in safer long haul trucking, it removes the danger the driver might doze off at the wheel during that long, boring stretch of road. Soon the systems will be good enough that the human in the truck could even sometimes put his seat back and take a nap while the system is in control. The system could alert the driver when a situation arises where a skilled human needs to take over.
The technology might not yet be prime time ready, but it is going to get there.
Stephen says
That kinda defeats the purpose of the autopilot if companies have to pay someone to be on the truck still, they are still gonna have to be licensed to drive one. Not to mention all those sensors and systems are going to be very expensive.
MrYowler says
They won’t pay him for the time that he isn’t driving. He will still be trusting his life to a computer, but they’ll pay him peanuts for the local haul time, and make him sign a waiver of liability for the risks that he is taking. Or, they’ll have a pool of local drivers at each end of the run, to do the hard part. If it works, there will be plenty of out-of-work drivers, desperate for some kind of income…
gail morra says
How do you know?
MrYowler says
I already work for a carrier that makes drivers sit for layovers up to five hours long. They are still responsible for the load; they can’t leave it unattended – but they don’t get paid for the time, even though the job pays by the hour.
If this is how compensation works during a “driver shortage”, don’t expect things to get better when there is a surplus of drivers, because of automation. Compensation has not improved with labor force surpluses in any other industry.
I know, because it is a repeating, consistent pattern, that is eminently predictable.
road62 says
if drivers were not already penalized for “taking a nap” , the supposed great danger of us falling asleep on the road would be greatly minimized. No company is, in the long haul, going to invest in both the robot and the driver. The cost is too great and we all know that money is the bottom line. Safety equates to money.
Look at the May 18th Texas 87.9 million judgement against
Werner, and the driver did no wrong. The other 101 million judgement
in Texas just a few days ago, for a company that knowingly hired
a driver with recent DUI’s, and drugs in his system at the accident.
The effect of the robot is to run non-stop. Already been accomplished,
California to Florida.
The trucking industry has never been about the driver or his/her safety.
It is about the broker/shipper/receiver and the insurance carriers and the bottom line. Otherwise, DOT and Congress would have put a stop
to load warehousing on the trailer, 10 hour holds in the delivery or pick up location, and added safe haven rules before the ones just announced for Personal Conveyance.
Been driving 41 years with no accidents,DUI or speeding tickets.
I have live and seen the carnage, the destruction of trucker families,
the very poor working conditions. So the ELD is not enough, even when it counts your day down to the second. The ELD and the various assorted tracking hardware, including use of your phone to map a location, are being data mined to program the A.I. to do your job.
You will be minimized to swinging the cargo door.
Tommy Molnar says
I think the money would be better spent repairing our aging highways, although a dozing driver will always be shaken awake by hitting some of these giant potholes.
We’re a long, long way from driverless cars, let alone trucks.
Stephen says
Exactly.
JACKSHEET says
REALLY! would have been nice to have 40 million dollars worth of parking along US highways main routes eh? And uh, wheres an autonomous truck gonna park when it breaks done, plan stupid really!
jerry says
Lidar, sonar, radar, camera’s, thanks Mr. Burnett & others for bombarding the American public with even more elctro-junk, above & beyond cell phones, WiFi, etc. Sure, manufactures have self-paid studies to debate these health hazards, but we’re just beginning to see the long term effects on our systems & overall health.
And of course, there’s always tech-hijacking possibilities, (just imagine “re-routing” that multi-million $ load of drugs, or other high-value cargo).
MrYowler says
The hacker in me wants to start jamming sonar and radar… Seriously, it’s going to be sooo easy for someone ti create freight and equipment damage conditions, and possibly kill people, and then disappear as if they were never there – because they weren’t…
That aside; it’s not really the sensor technology that is tge problem, here. We have excellent cameras, which see far better than human eyes. The problem is that these machines have no capacity to *perceive* – only *see*. When a tire blows out, will they understand what has happened where they can’t see it? Will they grasp the concept of lost traction? Will they recognize the hazard posed by an erratic driver next to them, or an animal on the side of the road, and respond appropriately? Do they understand a person standing in the road with a stop sign in their hand? Cones?
It’s the capacity to not just see, but also interpret what is seen, and make rational decisions about the appropriate next course of action, which is going to limit these vehicles. Because we don’t hold computers responsible for judgement, the judgement calls that they make (if they ever gain the capacity to make them) will not be based upon concern for human life and welfare, but liability concerns and productivity.
The only way this works, is with dedicated, automated-only routes. And that will probably be aerial drones, not trucks.
jerry says
Agreed MrYowler, this type of intelligence is going to be very difficult, if not impossible to mimic, imho. The ego-driven folks racing to be 1st to market on this tech are treading on some dangerous ground.
bobtrucks2204 says
You are correct MrYowler. About a year before I got out the company got these new trucks with the latest gizmo electronics. Driving late at night, relaxed, then the brakes slam on because the lidar thought the overpass ahead was too low. Talk about messin your pants. I didn’t know what was happening for a few seconds. Then I figured it out. It was time to go.
Ed says
Think about all the eletrical problems we already have with simple sensors that shuts the whole truck down on the side of the road…. now think of all the safety sensors that those trucks will have… most likely they will have some sort of safety system built in to shut the truck down too….
Jakk says
Nothing but a scam to get money! This is never going to happen in our lifetime. I said it before and I’ll say it again. NOT going to happen any time soon!
Trey says
I would not get into one of these things. As soon as a accident happens they will blame the driver
JOHN H O'NEILL says
He’s one step away from creating a …what’s it called agian….oh yea…..a train. Just add tracks and you got this dude….lol….tech people so laden with other peoples money that they have ZERO idea what to do with it so they come up with bs like this.