Artificial intelligence blooper videos are a much-needed respite

A recent video from Boston Dynamics showing their robot Atlas as a design assistant (posted below) is absolutely stunning and serves as another reminder of just how amazing artificial intelligence (AI) can be.

Just don’t miss Atlas’ blooper video (also posted below). So far, it’s easily surpassed the Final Cut with 11 million views versus 5.2 million views for the Atlas makes no mistakes video.

This is interesting and teaches me a couple of things: first, robot blooper roles could be the next big thing (sorry BattleBots), and second, people like to be reminded how imperfect robots are. It turns out that AI has a human side, even if a clumsy robot doesn’t quite appreciate the humorous bent of its shortcomings.

As for blooper roles of autonomous cars and trucks? Yes, not so funny. After all, few topics in trucking arouse as much interest and anxiety as artificial intelligence.

On the plus side, AI can make life on the road safer through collision avoidance systems, which typically analyze and react faster than human drivers to avoid potentially fatal and costly accidents.

[Related: Autonomous startup gets it first OEM partner]

AI can also prioritize huge streams of data from telematics devices, allowing fleet managers to address device issues faster before they lead to costly outages, and identify the key drivers for improving fuel economy and fleet safety.

Route optimization is another important AI application. AI can consider weather and traffic (both real-time and historical) to suggest more preferred routes to save time and money. AI is also increasingly being used to connect loads to carriers more efficiently.

AI has also been used in warehouses, where automated forklifts and other cargo handling equipment can move loads for transit more efficiently. CCJ Innovator NFI Industries reported a 50% increase in productivity at some of its operations thanks to its autonomous mobile robots (AMRs).

READ :  Communications Platform-as-a-Service (CPaaS) Global Market

Autonomous trucks show promise not only in terms of fuel savings and improved safety, but also in terms of reducing travel time as they do not require mandatory driver breaks. Attending a Torc Robotics autonomous truck event in November, Torc CEO Peter Schmidt beamed when he told reporters that one of their autonomous trucks arrived at its destination “four hours early” in a pilot project with Schneider.

So there is a lot of good news from the AI ​​frontline making its way into the complex and challenging world of truck driving. There is also still a lot of room for improvement.

No Coffee For You When I recently spoke to a veteran truck driver who, for months, has been driving with forward-facing cameras and in-cab sensors tracking his eye movements, AI has proven to be both a blessing and a curse highlighted (a challenging topic, that can stay here).

This longtime driver had to break the habit of drinking his morning coffee and other beverages while driving. The reason? Every time he looks away and grabs that drink and takes a sip – no matter how fast and efficient he’s gotten over the years – his driver scorecard will automatically score a hit.

A few slurps here and there added up to his very first write-up from his fleet manager. Now he no longer drinks or snacks on the go for fear that he will be texted. Thankfully, he doesn’t drive long-haul, so fighting the urge for snacks and drinks isn’t as bad as it could be. Also, he said he feels he has become a safer driver under a strict AI tracking system. Still, it annoys him that he can no longer drink his morning coffee on the go – and who wouldn’t?

[Related: Autonomous truck spotted in the wild]

I can definitely relate to AI’s annoying eye-tracking feature. When I recently tried out Ford’s BlueCruise self-driving system in their electric F-150 Lightning, I was showered with warnings for looking slightly turned. The system even tapped the brakes a few times to let me know I’d looked away too many times.

READ :  Google chases Amazon and Microsoft in cloud with AI technology

Do not get me wrong. BlueCruise was great in traffic jams, where it could easily handle the annoying stop-and-go traffic. All I had to do was stare straight ahead while BlueCruise handled the braking, acceleration and steering. In that sense, I was a backup driver who rode along.

I’m sure you’ve seen some self-propelled systems come under fire. Tesla tops the list. After years of federal investigations, the National Highway Traffic Safety Administration (NHTSA) announced a recall of 362,000 Tesla electric cars due to continued concerns about the company’s Full Self-Driving (FSD) beta software.

Multiple federal investigations have found Tesla’s self-driving capability to be responsible for fatal accidents. Tesla has consistently instructed drivers to always be ready to take control of the car when in self-driving mode. However, not everyone follows the instructions — shocking, I know — and instead lets the AI ​​do all the driving, resulting in fatal collisions and an NHTSA recall.

Tesla’s recall amounts to an over-the-air update to improve the FSD software, which got me thinking about hacking. Any system that accepts external data – and I don’t know of any system that doesn’t – is obviously at risk.

[Related: Watch Torc Robotics autonomous truck deliver fuel economy gain]

But perhaps technological advances over the years have reduced the risk of hacking? I asked this question to Ogden Stojanovski, Chief Operating Officer and co-founder of Pronto AI, which develops autonomous driving systems for on- and off-road commercial vehicles.

“Hacking should always be an issue,” he said. “Cybersecurity is a big, big part of what we do.”

Stojanovski said the threat of someone hacking into an autonomous truck, potentially using it as a destructive and lethal weapon, is “much less likely than someone hacking in and maybe stealing a whole bunch of valuable data.”

READ :  Introducing AutoML: Let's do a reality check

Also, the threat of hackers creating infrastructure connectivity by stopping autonomous trucks on a busy road is more likely, Stojanovski said, than “getting them to do crazy nasty things.”

Regardless of advances in fighting hacking, Stojanovski reminded, this old game of cat and mouse is not going away anytime soon.

“Cybersecurity is a kind of arms race,” he said. “You have to be on your toes. You have to keep getting better and better and keep up to date with the latest threats.”

And not all threats come from hacking. Slow down and stop in front of self driving truck to avoid collision. Now you have an unmanned truck that poses less of a risk to would-be thieves, Luddites or any other malcontent wanting to harm it to make a statement.

There are others out there as well whose fun ideas can quickly become wacky headlines that aren’t limited to Florida. Austin, Texas definitely springs to mind, where swarms of people recently took over the busy city streets to run around and drift at the intersection.

This chaotic scene is reminiscent of TC Boyle’s short story Asleep at the Wheel, an interesting and cautionary tale about life in autonomous fleet cars. Boyle, who knows human nature well, reveals a scene in which curious and daredevil children attempt to overtake autonomous cars, which turns out to be tragically stupid.

Of course, most cutting-edge technologies pique our interest and can involve some degree of risk and controversy. But when it’s put on wheels and self-propelled with passengers and/or cargo, it’s arrived at the weird and fascinating crossroads of science fiction and reality, where the results can be both rewarding and unfortunate.