Imagine a future Blade Runner-esque workplace in which human and robot co-workers work side by side without it seeming in the least bit remarkable. As it turns out, you don’t need to be much of a futurist to conceive of such a scenario: It’s been the day-to-day reality of factory and warehouse workers for decades.
The term “automation” was first coined in 1948 by Delmar S. Harder, an engineer and vice president at the Ford Motor Company to describe the handoff of particularly heavy, repetitive, and dangerous jobs to machines in industrial settings like factories. The first industrial robots and automated warehouses began appearing in the 1960s — and have only grown in number since then.
Not all jobs can be easily automated, however. Consider the job of forklift operator, for instance. On paper (or, well, screen), forklifts fall under the same remit as many previous industrial robots: Lifting and moving heavy goods from one place to another. But operating a forklift is far from easy or predictable. It requires significant awareness and decision-making on the part of the operator. Forklifts can weigh significantly more than the weight of a loaded passenger car, have to grapple with uneven weight distribution, turn with their back wheels, rather than their front ones, and can be difficult to stop. It’s no exaggeration to say that they can be tougher to drive than a road car.
Operating the forklift is the most dangerous thing to do in the warehouse.
A Union City, California, startup called Third Wave Automation believes it’s cracked the problem, though. And it’s helping to create the self-driving forklifts to prove it. Using expertise in fields ranging from robotics to computer vision, Third Wave has developed technology that promises to help revolutionize the warehouses of tomorrow. Emerging from stealth this month, the two-year-old startup has just announced a funding round of $15 million to help it do exactly that.
“[Forklift driving is] a dangerous job that requires certification, and has been historically very difficult to keep staffed,” Arshan Poursohi, CEO and co-founder of Third Wave Automation, who was previously a roboticist at Google and director of engineering at Toyota Research Institute, told Digital Trends. “Going by OSHA-reported injuries, operating the forklift is the most dangerous thing to do in the warehouse. We’ve heard from countless warehouse operators that it’s hard to keep enough qualified folks on staff to get the bandwidth from operations that modern supply chains demand.”
Poursohi said he was unable to provide a customer list, but that Third Wave is currently piloting its self-driving forklift technology. Using the insights gained from these tests, it will then create a commercial product that it can sell. Rather than building forklifts from the ground-up, Poursohi said that it has developed a software platform that, with the right sensors, can turn ordinary forklifts into self-driving models. This could be used to either retrofit existing fleets or licensed to forklift manufacturers, which could build the self-driving sensors into new production models.
He did not give exact details of exactly which sensors these include, but said that they are “things like lidars and cameras, similar to [autonomous vehicles].” Judging from Third Wave’s early demos, the robo-forklifts will be capable of identifying obstacles in warehouses or factories and then plotting a safe course to complete journeys. Once there, they will then be able to pick up and drop off pallets containing whichever goods need to be shifted about.
The human factor
It’s no mystery why Third Wave would be interested in shaking up the world of warehouse logistics. Compared to front page headline-generating technologies like self-driving cars or even food delivery robots, moving heavy things around a warehouse or “fulfillment center” might not be the world’s glamorous sector, but it’s certainly a valuable one — in all sorts of ways. There are an estimated 850,000 forklifts in the United States, and it’s a growing market that could be worth in excess of $13 billion by 2025. With tech-savvy retail giants like Amazon snapping up robotics companies it thinks could help further streamline its operations, it’s not hard to imagine a lucrative exit for the company which successfully automates this crucial job.
There are likely well over a million forklift operators worldwide whose jobs would be put at risk by a fully automated solution.
But what does it mean for the humans involved? The demand for forklift operators continues to rise in the manufacturing, transportation, and retail industries. Since there is, at present, no mainstream forklift automation technology widely used, that means there are likely well over a million forklift operators worldwide whose jobs would be put at risk by a fully automated solution.
This is not, by design, what Third Wave has created, though. As Poursohi said, the technology developed by his company still requires humans to be in the loop; it just doesn’t put them in harm’s way.
“The other technologies you’ll find out there take the stance of trying to constrain the use case enough that the existing tech can solve the problem, most the time,” he said. “Our shared autonomy approach inverts this and solves what the robot knows it can do already autonomously, while actively recognizing situations that it doesn’t have full confidence. In those cases, we have built our systems to be able to ask for guidance from an experienced forklift operator, and then to learn from that guidance so that it will be able to build the confidence to handle similar situations in the future.”
A bit like the technology developed by Swedish autonomous vehicle company Einride, which seeks to replace regular trucks with self-driving pods overseen remotely by truck drivers, Third Wave’s concept still keeps humans in the loop. It just means that one human can operate far more vehicles than they would if they had to physically drive it. The humans are there for edge cases in which the computer gets befuddled. In the case of warehouses, this means that more goods can move through the same buildings with the same number of staff. Perfect for an industry like e-commerce that’s growing at an almost unimaginable rate.
“We keep track of a metric we call ‘fan-out’ which is the number of trucks that one forklift operator can monitor,” Poursohi explained. “The operator is not actively paying attention to individual forklifts, but rather being asked for guidance from the forklifts on demand.”
What does this mean for the future?
Of course, this does mean that the number of forklift operator jobs won’t grow as quickly as other parts of, say, retail warehousing may expand. The technology also means that, at least in the long run, operators may be training their eventual replacements. As Poursohi said, “in short, the system gets better continually after it’s installed, and can adapt to changes that customers will naturally make in their warehouses after we leave.”
But as with the 60-year headstart that factories and warehouses have over other businesses when it comes to humans and robots working together, so too is this dilemma an increasingly common one. Right now, the divide between what robots and humans can do frequently comes down to strength vs. ingenuity. Robots carry out tasks such as predictable lifting and moving of objects. Humans do the rest. Amazon, for example, uses robots — originally developed by the Amazon-owned Boston company Kiva Systems — to bring racks of heavy shelves to human product pickers, thereby saving time. The human pickers then use their human speed and dexterity to grab the right products off the shelves and put them in boxes.
Still, this relationship isn’t fixed. Thanks to advances in robotics and machine learning, it’s shifting all the time. Robots are now able to carry out jobs that, just a few years ago, would have fallen under the “human” side of the job description banner. In the years to come, this balance will further tip as robots become more and more capable and confident about carrying out a growing number of tasks.
For now, though, forklift operators aren’t going anywhere — although their jobs could be about to change in a profound way. Welcome to the future of logistics in the 21st century.
- Three words: Robot. Talent. Agency.
- OpenAI’s GPT-3 algorithm is here, and it’s freakishly good at sounding human
- Amazon moving in on a Silicon Valley self-driving startup, report says
- Lyft’s driverless cars are back on the streets of California
- Amazon buys autonomous vehicle startup Zoox for more than $1 billion