Crux Systems Insights

The continued rise of autonomous vehicles

June 24, 2016 by Eric Klein |


Autonomous cars and trucks will change the way we drive and the shape of our infrastructure. There is certainly a question of when, though despite the prediction from Elon Musk, it won’t happen in two years. The real question that remains is how will this work.

I’ve been using Tesla’s autopilot now for a couple of months and I love it. It’s a huge improvement in car technology and a massive improvement in safety. Autopilot shines in its most effective mode, when the car is in heavy, stop-and-go traffic. That’s when you as a driver want to check out, and traffic needs cars that just stay in the lane and keep things smooth.

But at 70 mph, Tesla’s autopilot is at risk of being more dangerous than safe. At that speed in traffic, you can’t disengage. You must still be fully alert and ready to take over in a heartbeat. The car performs well at this speed, but shows in more than a few instances that there is a lot of work still to be done.

This is the real issue: it is impossible for an autonomous car or truck to be 100 percent reliable. It is impossible for developers to conceive and solve for every possible scenario, which is why a driver has to get involved when problems come up. In fact, with the autonomous Freightliner Inspiration Truck, the driver has to stay involved, sitting in the driver’s seat, and actively engaged (as they are still required by law to be). But remove the driver, or even legally allow the driver to stop paying attention, and a lot more questions come up that don’t have answers yet.

When I was working on automated guided vehicles and automated stacking cranes, regardless of whether we were in Rotterdam, Virginia, Antwerp or Brisbane, the answers to those exception questions were always the same: stop the equipment and get a person involved. Exactly how that occurred differed from site to site and from equipment vendor to equipment vendor. But there was an advantage in every situation that autonomous cars and trucks won’t have -- or don’t yet have. And that is a small, controlled area where the equipment operates and a standby force of technicians ready to react and deal with equipment issues.

How will companies deal with the autonomous car (with its passengers) that gets stuck on Highway 1 on its way to Stinson Beach? Or the autonomous truck that gets stuck in the middle of the 710 at 2 pm? I’m not saying these are deal-breaking issues. They are just issues that still need to be solved. And there are countless others like these.

The rise of automation is still coming and is closer than it has ever been before. But it is also a lot farther away than the investors in the VC world would like you to think.