California-based Solo Advanced Vehicle Technologies wants to build electric, purpose-built heavy-duty truck platforms for autonomous freight.
These platforms would be compatible with any autonomous driving software, so autonomous vehicle companies could simply layer on their self-driving software and sensor suite and be off. The startup, founded by Waymo, Tesla and BMW alums, aims to address the inefficiencies behind retrofitting existing human-centric trucks for autonomous driving.
“For about five years, I spent time doing sensor integration analysis on Class 8 trucks for Waymo and worked very closely with Paccar and Daimler, and it became very apparent that you have this really high-tech sensor suite and AI from the AV industry, and then you’ve got this ancient diesel human-driven truck that really hasn’t truly changed in close to 100 years,” Graham Doorley, founder and CEO of Solo AVT, told TechCrunch. “When you look at the problems of integrating autonomy on them, it’s just a plethora of issues.”
Issues like the inability to put sensors in optimal locations, inefficient power trains and compromised aerodynamics, Doorley said, noting the need for a hardware and software solution that can both handle advanced autonomous technology and help it commercialize at scale.
With its recent $7 million seed funding round led by Trucks VC with participation from Maniv Mobility and Wireframe Ventures, Solo AVT is gearing up to finalize the design and build of its first test vehicle, a battery-electric Class 8 truck, and begin testing this summer, the company said. Solo AVT will then use the knowledge from its “test mule” to inform the design and engineering of its alpha truck, the SD1 Heavy. This truck will feature stabilized and optimized placement of sensors that the company said can’t be done with legacy trucks.
The fresh injection of capital will also help Solo AVT grow its team, specifically with engineers, from eight now to about 18 by the end of next month.
Solo AVT’s trucking platform will not be human-drivable and therefore will have no human controls, like a driver’s seat or steering wheel. They will, however, be built with a teleoperations-based system on top of the AV stack so that humans can still be involved in maneuvering around hubs or to take over if needed, Doorley said, noting that the company is expecting to have alpha vehicles on the road by 2024.
Swedish freight technology company Einride is also building an autonomous electric solution to the trucking industry. The company just hired its first “pod operator,” a trained and licensed truck driver to remotely monitor operations of Einride’s autonomous pods, and last year began testing some of its tech and electric trucks in the U.S.
Autonomous trucking companies like Waymo Via, Aurora, TuSimple and Kodiak Robotics all have partnerships with traditional OEMs that often began with retrofitting trucks for testing, but have evolved into creating purpose-built trucks together. Waymo Via, for example, is developing a unique version of Daimler Truck’s Freightliner Cascadia “from the ground up” specifically for the Waymo Driver. TuSimple is working with Navistar to build purpose-built autonomous trucks, and Aurora is working with Volvo VNL.
While not technically retrofitted vehicles, these trucks will still be built with the look and feel of a Class 8 truck — they’ll run off a diesel engine and they’ll have manual controls. That last has been something of a requirement for certain companies to enable testing, particularly in states like California, where companies usually begin with drivered testing permits and can graduate to driverless.
“We envision there will be certain operations that still require manual driving, for example, when a truck is in a loading yard,” Cheng Lu, former president and CEO and current adviser to the CEO at TuSimple, told TechCrunch. “It’ll be diesel engines given EV or hydrogen for long-haul trucking still have certain limitations, but of course, we are hopeful of alternative powertrains and look forward to adopting them.”
Legacy OEMs bring a lot to the table, including being able to provide warranties, after-market support and spare parts, said Lu, noting that being able to remotely control an autonomous truck wouldn’t change how TuSimple thinks about it.
“Long term, though, there could definitely be interesting design changes we’d explore,” a Waymo spokesperson told TechCrunch. “We do expect to see changes depending on how our technology (and technology more broadly) evolves, as well as how the use cases of autonomous trucks evolve, since these both play a very important part in how we design our vehicles.”
Fully autonomous trucks need to be built with redundant systems that can take over if a control fails. The autonomy system will also need to be plugged into sensory inputs and actuation controls. It’s a complex task, one that AV companies like Aurora have turned to legacy OEMs for in part because they trust OEMs to manufacture trucks that are safe and can be scaled, but also potentially because there hasn’t been much else on the market.
Aside from testing its mule this year, Solo AVT is planning on following up its seed fund with a Series A raise in the coming weeks. Doorley wouldn’t say how much the company is hoping to raise, but he did say the number would be significant and that Solo is shopping for strategic investors that can help move the business forward – like shippers, carriers, manufacturers, AV companies and suppliers. And no wonder.
Although there is a strong software element involved in Solo’s business, at the end of the day, it is also an electric vehicle company. And as we’ve seen from the likes of Rivian, Lucid Motors and Lordstown Motors, sticking to production goals is tough in this economy, even if you’ve got all the venture capital money you could possibly spend, and so far, Solo doesn’t have much cash to burn.
To this, Doorley responded that passenger vehicles and pickup trucks already require far more parts, tooling and volume than a Class 8 truck would. Add on the fact that there’s no interior to design for Solo’s vehicles, nor is there a complex body shape to stamp.
“We’re sort of banking that there’s lower capacity and lower volume requirements for these,” said Doorley. “We’re also a couple of years away, so hopefully the supply chain issues of today will be somewhat more resolved.”