This is a question For you: How seriously should we take Amazon’s toy home robots? Perhaps the best way to frame it is: When do we take Amazon home robot play seriously? I realize these seem like pointy questions, and I must specify that they are not really Amazon specific. They are more a result of burning in the past.
The road to a home robot is full of good intentions from companies large (Sony) and small (Anki, etc.). For decades, robots have been a kind of industry shorthand for future innovation. Do you want the world (and most importantly your stakeholders) to know that you are focused on the future? Roll out a bot at your press conference, and who cares if it ever comes out?
Obviously, Amazon has addressed that last bit of potential criticism. Astro came out. The company, which was announced a year ago this week, launched the robot as part of its “Day One Edition” program, offering it with limited availability and a hefty price tag ($1,500). During a call before yesterday’s Alexa event, Ken Washington, the company’s head of consumer robotics, was upset when she explained that Astro’s rollout found the company testing the waters of robot driving.
“[Day One] It’s a way for us to get these products into their hands quickly,” the executive told me. “It’s not for us — not knowing their interests — knowing what they want to see more to add to it. … We have received hundreds of thousands of incoming orders and continue to manufacture and publish Astros for those customers who place those orders. I can’t share the actual sales amount because we don’t share data. But we got off to a great start.”
To some extent, I think the rejection has to do with framing. I see turning on Astro early on was an attempt to determine if people eventually wanted this way of a home robot. Amazon appears to be taking this desire as a kind of premise and instead looking at the software as a way to accurately determine why people want Astro. With the announcement that a new SDK will be introduced to the University of Michigan, Georgia Tech, and the University of Maryland this year, it’s probably time to start thinking of Astro as a platform first.
There is a feeling that the company is almost behind the iRobot model. The company identified a need (sweeping) and built a robot for that purpose. iRobot tends to be very deliberate in its approach. That’s why the company has taken so long to offer a convenient 2-in-1 cleaning/wiping robot.
“The customer is very excited about the convenience of the two-function robot, so we needed to build one,” CEO Colin Angel told TechCrunch. “But, being an iRobot, we needed to build one, rather than do it in a way that didn’t deliver. Right now, most two-in-one robots are really one plus one.”
Of course, assuming the FTC didn’t put the kibosh in the deal (not a foregone conclusion yet), the two are about to be part of one big, happy family. If iRobot ends up turning into Amazon’s line of consumer robots, it will be impossible not to take the company’s ambitions seriously. The way the company has accelerated the categories of warehouse and fulfillment bots is certainly a clear indication of what the company can do with essentially unlimited resources. And while it has already built its own home robot, it’s easy to imagine Roomba working with the same foundational ability as Kiva to play consumer robots, going forward.
We’ve got a whole bunch of news to turn to this week. I’ll start things off cheerfully. In July 2021 we covered the news that Cassie from Agility Robotics ran 5km. It was a great achievement and a nice feather in the OSU cover. Of course, the company has become better known by Cassie’s follow-up, Digit, but the original ostrich-inspired robot is still involved in some interesting stunt work at a few universities.
Jonathan Hurst, chief technology officer at Agility, told TechCrunch:
Many dynamic behaviors are difficult to mathematically represent, especially any physical interaction such as walking or running. Machine learning techniques have the ability to represent this complexity, but so far they have struggled to find good solutions or to translate from a simulation to a real machine. We learn how to use our experience and knowledge of legged locomotion to guide machine learning, and get results that outperform other technologies. That’s interesting! No free lunch – The machine learning system probably won’t detect useful new behaviors on its own; We have to understand the goals and direct them well.
The Amazon-backed startup announced this week that the robot has set a world record, Running 100 meters in 24.73 seconds (It’s a feat the company teased during a panel discussion at a Robotics event in July.) That’s still way off Usain Bolt’s 9.58 seconds, but it’s still a very impressive foot-up for a bipedal robot (there’s also something to be said for knowing we can still get past them a little longer).
Big increase for Avidbots this week. The Canadian company has announced a $70 million Series C for its industrial floor cleaning robots. The round was chaired by Jeneration Capital and included True Ventures, Next47, SOSV, GGV Capital, BDC Capital, Golden Ventures and Kensington Capital. This brings the company’s total increase to $107 million.
Avidbots plans to add an additional 100 people in products, engineering, sales and marketing over the next year. CEO Faizan Sheikh says:
We are very excited about the future as this funding allows us to accelerate our schedules to bring new products to market as well as continuously improve our autonomous driving software and services for existing customers to deliver a better experience and greater value.
Livin Farms is one of the most interesting bot users I’ve come across in recent weeks. The Austrian company raised $5.8 million for the Hive Pro System, which raises black soldier fly larvae to produce a sustainable protein powder.
“[O]“Our customers are making a huge contribution to fixing broken diets and thus saving the planet,” said Katharina Unger, CEO, commenting to TechCrunch. The process takes about 11 days, at which point the larvae become “half a ton of biomass plus half a ton of fertilizer.”
Natasha has some interesting news from Europe this week. The European Union recently updated its liability laws to include artificial intelligence. The new AI liability directive will make it easier to prosecute systems such as drones, robots, and smart devices.
“The principle is simple,” Justice Commissioner Didier Reynders told TechCrunch. “The new rules apply when a product powered by artificial intelligence technology causes harm and that harm is the result of an error made by the manufacturers, developers or users of the technology.”
Meanwhile, Rita has written about an initiative aimed at breaking China’s reliance on US-designed semiconductors. The motive here is easy to understand, given how much US sanctions during the Trump administration have sidelined Huawei in the past few years. Horizon Robotics is leading the way with a whopping $3.4 billion in funding to date.
And finally, we’re getting ready for Tesla’s AI Day. That event is tomorrow, and I plan to stay up late to help Kirsten cover it — and peek at the company’s longtime Optimus bot. Let’s just say my expectations have been tempered enough, but I’m open to pleasant surprise.
Ready, set … Subscribe to the launcher.
#sprint #marathon #TechCrunch