It’s been a few years since I paid a visit to iRobot’s headquarters. That was pre-pandemic, pre-Amazon, pre-newish CEO. There’s been turnover and downsizing, too, but that’s a bigger corporate conversation for some future newsletter. This Thursday morning, I just want to focus on the office lobby – or at least what used to be in the office lobby.
(I’m sure some current employee will let me know either way.)
The iRobot museum offered a crash course on the company’s early days – and provided a good story intro for a tech journalist in a pinch. Mars rovers, baby dolls, military robots – there’s no particular throughline, beyond a young robotics company trying to make it work.
Cofounder Helen Grenier ultimately convinced VCs to take a shot on the firm, but it had already managed to eke by for some time on grants and grit. iRobot’s early days epitomize the story of the scrappy startup whose fortunes could have just as easily gone the other way.
“It’s unlikely anyone would have deliberately chosen the path we followed,” Joseph Jones notes in his new book, Dancing With Roomba. “But after several years of wandering in the robotic wilderness, we had accumulated some key advantages in business, technology, and public relations that would greatly benefit us when we took on Roomba.”
Jones and I will discuss the advantages – and to what degree any of them are truly applicable to future startups – in an upcoming newsletter interview. Twenty-five years later, Roomba and its ilk still feel like a brilliant fluke as far as home robots go. And here we are, still waiting for that next big follow-up to arrive (more on that below, actually).
The first bit about not deliberately choosing that path certainly resonates deeply with many of the conversations I've had recently around Automated. Success, as we wrote a few weeks back, is not a fixed point. The world around us is ever-changing, and canny founders will allow their companies and products to change along with it.
At seven, Aadeel Akhtar visited Karachi, Pakistan’s largest city, and his parents’ hometown. “That was the first time I met someone missing a limb,” he explains. “She was my age, missing her right leg and using a tree branch as a crutch. And that's what inspired me to want to work in this field in particular, in making limbs for people and now robots.”
Such journeys are, of course, never so straightforward. Life has its ways of throwing up roadblocks for even the most charmed among us — even those who do manage to devote our adult lives to fulfilling a childhood dream dreamt halfway around the world.
The story stayed with Akhtar, through multiple cross-disciplinary degrees — including a Ph.D. in Neuroscience an M.S. in Electrical & Computer Engineering — culminating in his team’s research at the University of Illinois at Urbana-Champaign.
“The idea would be that I would work at an academic hospital and have my own lab, like an academic lab at the hospital where I'd see patients once a week, people with limb differences. And then the rest of the week, would build devices like this,” Akhtar holds up the latest version of Psyonic’s Ability Hand, during our conversation on this week’s podcast, “where we would test them on potential patients.”
Development was a big piece of last week’s Blue Jay unveil. The move was a kind of strategic shift for Amazon Robotics, which has tended to spend the majority of product reveals focused on the technology itself. Here, however, chief technologist Tye Brady devoted a chunk of his stage time not just to how the industrial system was created — but how quickly the team was able to do so.
Specifically, the relatively compact, six-arm system was iterated on in “just over a year.” Quoting directly from Amazon’s copy here,
Blue Jay’s development moved from concept to production in just over a year — a process that formerly took three or more years for earlier Amazon systems like Robin, Cardinal, or Sparrow. The reason: Years of trial-and-error were condensed into months of development thanks to advancements in AI.
It’s an interesting note, to be sure. Anecdotally, I do feel as though robotics firms these days are leading with speed of development as a key selling point. This is particularly the case among emerging humanoid firms — a technology (along with AGI) — many parties insist is much, much further from fruition than Silicon Valley would have you believe.
What are we up to, like one GTC a month at this point? Seems like every time I open Instagram, there’s a fresh photo of CEO Jensen Huang fist-bumping a couple of humanoid robots. And who can blame the guy? NVIDIA casually waltzed into a $5 trillion valuation yesterday. The dude can wear all the leather jackets at all the developer conferences with all the robots he wants. What NVIDIA really leaned into during the one hour, 45-minute-long keynote, however, was patriotism – the event was, after all, streamed directly from the nation’s capital. The show kicked off with a video tribute to the U.S., before Huang told the crowd, “It’s hard not to be sentimental and proud of America, I’ll tell you that.”
The CEO then devoted much of the talk to “reindustrializing” the country through robotics, physical AI, and Omniverse digital twins. Relatedly, Agility Robotics, Amazon Robotics, Figure, and Skild AI are the first partners to use NVIDIA's three computer platform to build and deploy fleets. Also relevant to our newsletter rounding up needs is IGX Thor, an enterprise-focused robotics processor designed for running generative AI models on the edge. The IGX Orin successor is focused specifically on industrial, robotics, and medical/healthcare applications, with launch partners including, Diligent Robotics, EndoQuest Robotics, Hitachi Rail, Joby Aviation, Maven, and SETI.
Speaking of, this is the next version of Diligent Robotics' healthcare robot. Moxi 2.0 arrives on the heels on 1.25 million deliveries performed by its predecessor, constituting what the Austin firm calls “one of the largest datasets of real-world human-robot interaction.” The earlier system is currently deployed in 25 hospitals, largely focused on delivering medications and samples. The hardware itself looks similar to the original Moxi at first glance, including the wheeled base and a mobile manipulator that can be used to press elevator buttons, so the robot can more freely roam the hospital without direct human intervention.
The company notes that Moxi 2.0 was built with scalability in mind, writing, “The new design is optimized for manufacturability and hardened to support the fleet’s quickly growing scale." Diligent adds that the new version allows hospitals to now deploy 15 units per site. “As a result, Diligent expects to double its hospital footprint annually, and deploy thousands of robots deployed 2030.” This news comes arrives shortly after the firm announced its participation in AARP’s accelerator program, which will expand Moxi’s footprint into non-hospital care facilities.
1X has opened its NEO home robot up for pre-order with U.S. deliveries set for some point in 2026, according to the company. A (refundable) $200 deposit is due upon sign up. The RaaS subscription will run you $500 a month, or you can just straight up own the robot for $20,000. This story – understandably – received a lot of mainstream press coverage. Between 1X putting giving itself a (albeit broad) deadline, Figure pivoting much of its focus to domestic tasks, and Tesla promising a chicken in every pot and Optimus in every garage, it's a great time to restart that conversation about precisely why the home has been such a tough nut to crack, and how companies like 1X plan to center safety in their roll out. Working alongside a robot on a factory floor is one thing. Inviting it into the home to help care for your parents can and should meet a significantly higher threshold. Tele-op learning in the home has raised eyebrows, as well.
Back in May, Kiwibot CEO Felipe Chavez told me the company had paid a lofty – but undisclosed – sum to purchase Robot.com. The move was the beginning of a big rebranding effort. Kiwi, which cut its teeth making last mile food deliveries on college campuses, had already begun to diversify, adding a warehouse robot to its fleet and purchasing an ad firm to essentially turn its systems into rolling billboards. On Wednesday, the company announced that it has fully transitioned to the new name and site. Robot.com, which now has around 500 robots operating in the real world, also launched a pair of new systems. There’s R-Dog a quadruped designed specifically for outdoor advertising, and R-Noid, a wheeled humanoid for the home. Those systems join the logistics robot, Cargo, and existing delivery robot, which maintains the Kiwi name.
Last week’s Delivering the Future event was sandwiched between a rough pair of news items for Amazon. On the eve of the event, a leaked internal memo suggested that automation could reduce future hiring needs by up to 600,000. Amazon called the documents “incomplete and misleading,” noting that it is “actively hiring” across the country. Earlier this week, Amazon confirmed an unrelated 14,000-person corporate layoff, in an open letter titled “Staying nimble and continuing to strengthen our organizations.” Amazon said the massive reduction is about "reducing bureaucracy, removing layers, and shifting resources to ensure we’re investing in our biggest bets."
Intrinsic is opening up access to its vision model (IVM) as part of a new developer challenge in partnership with Open Robotics. The system integrates directly with CAD models for “sub-millimeter accuracy” without training on new parts, per the Alphabet spinoff. Announced at this week’s ROSCon in Singapore, the new AI for Industry Challenge asks devs to solve “a similar problem that’s representative of real-world factory use cases, including tasks that increase in complexity from cable insertion to cable management.” They’ll have to train an AI model to manipulate electrical parts using open-source tools, like Open Robotics’ Gazebo sim, DeepMind MuJoCo or NVIDIA Isaac. The challenge officially kicks off in February.
Aadeel Ahktar (Psyonic)-Psyonic CEO Aadeel Ahktar discusses the company’s journey from human prosthesis to humanoid manipulation.
Tye Brady (Amazon)- Join Amazon Robotics' chief technologist on a journey from undergrad rockets to rolling out one million robots.
Brad Porter (Cobot) - Longtime Amazon VP Brad Porter knows what it takes to deploy robots at scale. He's now doing the same with his own company, Cobot.
The Association for Advancing Automation (A3) is North America’s largest automation trade association representing more than 1,400 organizations involved in robotics, artificial intelligence, machine vision & imaging, motion control & motors, and related automation technologies.
.jpg?upscale=true&width=1120&upscale=true&name=Untitled%20design%20(2).jpg)
-1.jpeg?upscale=true&width=1120&upscale=true&name=download%20(1)-1.jpeg)
-1.png?upscale=true&width=420&upscale=true&name=image%20(1)-1.png)
