In this weekly series, CNBC looks at the companies that made the first Disruptor 50 list 10 years later.
One of the lasting lessons of the iPhone era is that Steve Jobs ushered in something consumers didn’t know they wanted until they saw it. Robotics pioneer Rodney Brooks has worked on similar ideas, with varying degrees of success.
His robot vacuum cleaner Roomba and its parent company i robothas more or less made the robot vacuum category a cheap and must-have item for many consumers, but was recently sold to Amazon for $1.7 billion.
“iRobot found a product and market fit. It had a feature set and cost that many people were willing to pay for, and a pack that cleaned the floors very well.” Engineering Management, University of California, Santa Barbara program, and robotics expert.
But Brooks’ other big early idea didn’t pan out the way he or the world envisioned.
Rethink Robotics was founded in 2008. Cobots — robots that work alongside humans in ways that traditional automation couldn’t (i.e., are far less dangerous to humans), and without the fences most industrial robots keep inside — is a market waiting to take place among industry customers. But despite the novel ideas Brooks and his Rethink colleagues brought to the table, that first cobot was never well received.
Indeed, it was circulated in the press. If you’ve been following tech news at the beginning of the last decade, you’ve seen Baxter and “his eyes” staring out of the screen between his arms at one point in the photo or video footage. On CNBC, in 2013, he learned firsthand the limitations of working with new industrial robots when he invited Baxter to a live segment during Rethink’s inaugural inclusion on his Disruptor 50 list. Getting Baxter on his TV studio set was harder than he imagined.
He may no longer be a part of the robot future, but no cobot history can be written without Baxter.
“There is no cobot without Rod Brooks,” Bean said. “Rethinking started the industry”
After much early fanfare, Rethink’s reality has set in. By 2018, it was struggling to expand its business and find enough buyers for Baxter. His two-arm design proved to be a novel idea, but it was a mistake. “Nobody needs a dual-arm robot,” Bean said, explaining its design decision as “a human projecting his physical form onto the robot.”
It turns out that Rethink moved too late to a one-armed robot named Sawyer. However, there was another problem with this technique.
Rethink your approach with elastic actuators, a technology that one of the Brooks co-founders specialized in. This allowed the robot to perform “force sensing”. This is an approach favored by the company because it makes robots safer around their human colleagues. Rethink’s design also makes robots less expensive and less reliant on industry-standard motors and related components.
Mechanical engineer Paul Maeder, an early investor in Rethink through a venture capital firm, told the American Society of Mechanical Engineers: Post-mortem analysis on rethinking It went deep into technical shortcomings, and cheaper parts and force sensing seemed like a way to lower prices in the robotics market and appeal to customers.
And that brought the price down.
“The reality is that their prices were very low, surprisingly low,” Beane said. “They have done an amazing job of sourcing and designing materials, and have come up to a third or even a quarter of his price on other robots.”
However, Rethink never achieved the market penetration or scale necessary to sustain its business. As its financial situation deteriorated, Rethink became an acquisition target for a Chinese company. According to Beane, the deal would have been a good opportunity to scale the company. However, according to the company, the deal was scrapped at the “last minute.” 2018 bankruptcies followed. Rethink was acquired out of bankruptcy by German automation company Hahn Group. Hahn Group is still working on this technology today.
In a statement provided to TechCrunch at the time of service termination, Rethink said:
It’s true that Rethink had no shortage of interesting and innovative ideas, even if they weren’t suitable for the market.
According to Beane, one of the most interesting things was that workers could program the cobots. “People don’t use this feature much anymore, but it’s a graphic user interface system that doesn’t require code and can be taught and trained to do the work of a worker,” Bean said. Told. “But no manager turned on those features,” he added.
This is an idea Brooks is still working on through his latest robotics startup Robust.ai, which he co-founded with cognitive scientist Gary Marcus. Carter building a warehouse cobot, which is like a mobile shopping cart that provides transportation within a fulfillment center. “It can be driven in any direction, it can be worker programmable, and it can physically collaborate,” he said. “We have a lot of the same DNA”
With Sawyer becoming the leading cobot, Rethink invested in technology to integrate cobots with existing industrial automation. While this was an interesting approach, it was ultimately another bottleneck that cost engineering time to connect and communicate with machines such as conveyors. “Even just changing the conveyor speed is very costly,” says Bean.
The DNA that became standard in the cobot market that Baxter, and later Sawyer, hoped to dominate, came from Rethink rival Denmark-based Universal Robots. In Bean’s view, the company’s cobots may “look very boring,” but they were what the market ultimately wanted.
“In the end, a series of elastic actuators probably wasn’t the best idea in the world,” Maeder told ASME. “What customers really want is a low-cost, simple, fast, repeatable robot. They want to place something in this exact spot over and over again.” Ultimately, achieving that was a lot more complicated than it was for us, as some of our competitors weren’t even trying to sense force.”
Even with the two arms Baxter had, buying two single-armed robots could always be bought if a buyer really wanted that approach. That’s Universal Robots. Teradyne 2015 — UR3, UR5, and UR7 cobots lead to increased sales and excel. Successive Generations of Collaborative Robot Lines It continues to hit the market.
The opportunity for robotics technology remains significant, but it still lags other automation approaches in market penetration.Robot sales in the North American market Growingand the Teradyne unit, led by Universal Robots, is watching We are also seeing steady, if not explosive, sales growth. Revenue was $300 million in 2019, but increased to $376 million last year after the Covid plunge. According to Wall Street estimates, this year’s sales he could reach $ 440 million, a growth of about 18%. Its growth rate is higher than the percentage of overall revenue it represents, still less than 15%.
“There’s going to be more,” Bean said. “Progress often seems slow…but the usefulness of robotic systems that can perform common tasks at a reasonable cost is extraordinary and worth billions, perhaps trillions.” he added.
And costs continue to drop, from batteries to sensors to software. In other words, the price/performance ratio of robots is steadily increasing. But where cobots find their greatest utility is an open question. The first decades of industrial use were often a marketing ploy or, at best, an experiment rather than a proof of widespread adoption, but based on an aging demographic, cobots There are reasons to anticipate multiple roles.
“Anyone who says otherwise in my lifetime will need help, so hope it happens in your lifetime.
He notes that jobs outside the manufacturing sector that cobots are associated with are increasingly being used, from warehousing to retail to healthcare (imagine cobots delivering supplies to nurses) to retirement communities. We anticipate that this will be an area where “Manufacturing is about high throughput and high consistency and can be automated without cobots,” Bean said. “We’re just seeing this feature being useful and being used on a large scale.”
Rodney Brooks, one of the robotics experts, talks about an aging world and a shrinking workforce.he wrote Blog post covering his annual predictions “Soon the elderly’s homes will be cluttered with too many robots.”
Wall Street analysts are eyeing the opportunities posed by chronic labor shortages and related changes taking place in the global outsourcing paradigm on which economies, including the United States, have relied for decades. “Near-shoring” and the on-shoring of more manufacturing activities are increasing labor demand in a tight labor market. One answer is automation. Technologies like Universal Robots are relatively easy to program and implement.
But one of the big problems cobots haven’t solved yet is the same problem that Brooks started rethinking. That means understanding what everyone wants from this technology, the “killer app,” so to speak, from cobots. Universal Robots has a variety of applications for its technology, but no single application that creates strong demand in one very high volume market. This is a solvable issue, but still a work in progress. A robot running a popcorn station in a movie theater or a barista in his cafe is not a career opportunity that puts cobots at the heart of the economy. But people are in short supply, and automation must be part of the solution, even if it doesn’t happen by tomorrow.
The world of true physical collaboration between human workers and cobots is yet to come, says Bean. “We’re at odds with each other,” but “we’re getting there,” he added.
The most widely deployed automated systems remain the ones that are risky and shunned.
However, Beane thinks Rethink has come closer to solving the problem than he is credited with, something that has yet to be proven in his new life under Hahn. Hahn didn’t respond to a request for comment by press time.
“In another eight months, we might have had product market fit. It was cheap and reliable. I really believe we saw the iPhone moment,” Bean said. rice field.
On his personal blog, Brooks summarizes the “rethink” story: Since the idea was first developed in the Stanford AI Lab in the early ’70s, it has finally been freed from using a computer-like language to control it, as all robots have had. . We still have a lot of work to do. “