New Open-Source Software Looks To Kickstart The Autonomous Revolution

By now, you’ve no doubt seen the headlines positioning Amazon and Tesla at the forefront of fully-autonomous vehicles and transportation. Whether you want to call it a war between two tech giants or an ego-fueled squabble between billionaires, it’s a race (with no drivers, of course) to bring a safe and viable fully-autonomous driving solution to the real world. There is, however, another approach: collaboration.

That’s the tactic of Project ASLAN, which officially launched earlier this month. ASLAN is a supergroup of tech companies, research groups and universities, focused on bringing a fiscally practical, fully-autonomous vehicle solution to public roads. I sat down with one of the founding members of the group, Street Drone CEO Mike Potts to find out why he thinks a collaborative effort and open-source material is the only way forward.

Project ASLAN isn’t some big tech company’s pipe dream fueled solely by marketing hype and headline-grabbing claims. The company is primed and ready to ship out its specially-equipped autonomous-ready vehicles and the software needed to guide that hardware. It’s a landmark occasion and one, Potts admits, could only have been done with the collective brainpower of ASLAN group.

“There were an awful lot of companies and universities who were struggling with the automotive part of autonomous testing. There’s a lot of great software engineers out there, but cars and doing proper automotive testing is pretty tough. And that’s how Street Drone was born. We created a package people can plug into an autonomous-ready vehicle,” said the Street Drone CEO.

Street Drone specialized in the hardware part of the business, outfitting all-electric Renault Twizys and Nissan NV200 cargo vans with the necessary components for autonomous driving. The cockpits of vehicles look no different than factory vehicles, aside from the extra screens. The extra screens are for the engineer and a safety driver, which each vehicle needs. Potts admitted though, the two-person Twizzy is “a tight squeeze, but everything is still hidden.”

How hidden? Well, the steering column is replaced with a column containing two motors, one of which operates the normal power steering and one that is operated by the ASLAN system. The brake gets pulled by a little pulley. “You can’t see the hardware and it doesn’t impact normal driving of the vehicle if the safety driver needs to take over,” according to Potts.

For the software and cloud connectivity, the Japanese group AutoWare was brought into the fold. Most importantly, the software is open-source, meaning it’s free for anyone to use. A concept some, including Potts, believe will help drive the competition and the industry forward.

Easy integration was just one of the goals of the ASLAN group. Potts was adamant the number one hindrance holding the autonomous revolution back is cost. “The biggest problem with autonomy is it’s way too expensive. Waymo raised around $3,000,000,000 but probably spent a lot more with 600 vehicles on the road. That’s works out to about $5,000,000 a car. That’s a pretty expensive taxi.”

Potts continued, “That’s the reason for collaboration. It’s ultimately about the cost. You’ve got all these companies hellbent on creating the same system independently of each other. Some are coming together — Ford and VW have an agreement with Argo. But often, these companies are not working in parallel and not working collaboratively and it’s causing huge problems because there’s an awful lot of money being spent on solving the same problem again, and again.”

However, while Tesla and Amazon are focusing on passenger and shipping vehicles capable of driving on the highway, Potts says the revolution won’t start with speed. “There’s something specific that needs to take off first, and that’s not cars and robo-taxis. It’s slow speed stuff: last-mile deliveries, taking medicines from a pharmacy to a customer, etct.’ Things that trundle around the streets at slow speeds with no need for over complicated tech.”

ASLAN’s ideal customers would be shipping companies like UPS, DHL and even Amazon to a degree. “There’s no reason this system can’t work for mining, agriculture, or other low-speed environments, but we’re focused on street driving and deliveries for now.” Again, it comes down to cost. For ASLAN’s autonomous vehicles to make sense to any customer, such as grocery store chains or delivery services making local drop-offs, the autonomous system not only needs to work just as well as the traditional human driver, it also needs to be cheaper.

Potts broke it down what the cost per vehicle looks like for one of those potential customers using delivery drivers. “Right now, traditional trucks are about £70,000-£80,000, but with a driver on 16-hour shifts, over the course of a year are probably more like £150,000.” As of right now, (safe) fully-autonomous vehicles require an engineer present, to keep a watchful eye over the hardware. The next step for ASLAN is refining the tech so a human presence isn’t needed in the actual vehicle.

“The reality of the industry is you have to have a safety driver in play all of the time. Those guys are more like £150,000 each because they’re qualified engineers, not just delivery people. The cost of the vehicle itself, packaged with a computer, LIDAR, cameras, etc. means there’s no way autonomy can happen for a company like Sainsbury’s, unless you can get the cost down to a comparable point.” Potts continued, “If there’s one thing autonomous services enable, once you take the cost of the driver out, you can reduce the size of the vehicle. Big trucks, buses and tractors are big because of the space needed for the driver.” Once you can take the driver out of the equation altogether, lowering the cost of the truck itself becomes more of a reality.

That “reality” isn’t far off either. The past few months threw a sizeable wrench in the works, slowing any advancement in the industry to a halt, but Potts seems confident the open-source software tactic will help spur momentum when normalcy returns. Project ASLAN recruited teams of talented engineers from around the world to kickstart a more pragmatic approach to the autonomous driving conundrum. The hope is sharing their findings and progress with the rest of the world will take the baton further. Competition in the industry is healthy, but so is looking at the bigger picture, letting the ego go and fanning the flames of innovation.

Source: www.forbes.com