I told management I wasn’t coming back to work Tuesday if the Automated YouTube page didn’t hit 1K subs over the long weekend. So, thanks for everyone who helped make that happen — the subs, not the coming back to work bit. And extra double shout-out to the person who commented this on yesterday’s interview with Professor Daniela Rus:
Is that too ambitious of a goal to set for my return to work next week? Would it help if I told you our upcoming episodes are: Amazon Alexa VP Daniel Rausch, Generalist AI CTO Andy Barry, director of product for Atlas at Boston Dynamics Aya Durbin, Locus Robotics CEO Rick Faulk, Russ Tedrake, and MIT’s Yoel Fink?
I can keep going. How about Dr. Ayanna Howard, Clara Vu, Samantha Johnson, Jagdeep Singh, Kathryn Zealand, Andrei Danescu, Rajat Bhageria, Inna Lobel, and Karen Panetta? How about our upcoming interview with FieldAI’s Ali Agha at Automate next month?
That will be the third-ever live episode of the show, following GM’s Mikell Taylor at last year’s A3 Forum, and the aforementioned conversation with Prof. Rus, which we recorded in front of an audience at MassRobotics back in March. It’s a good one, I promise. We discuss the discovery of a new sperm whale alphabet.
Clearly the entire intro so far has been one long excuse to plug a bunch of upcoming podcast episodes. But hey, what’s another 999,000 subs when you’re working with an editorial calendar like that?
As I mentioned last week when I covered her piece on reframing the “Dull, Dirty, Dangerous” conversation, we’ve also recorded an episode with the RAI Institute’s Kate Darling. You’ll find a feature based on that conversation below. We discuss robot opinions. My take is that robot opinions are like iPhones. Everyone who doesn't have one is probably an android.
It’s also the reason I can justify two terrible Pleo puns in this week’s newsletter. I also have a soft spot for that little robot dinosaur. It’s a life goal of mine to start a sanctuary upstate for wayward social robots. Me, Pleo, Jibo, Keepon, Cozmo, Nabaztag, Mykie, Kuri, and a few acres of land for them to roam free.
Meantime, we’ve got some great newsletter content lined up. I just did another interview with aforementioned upcoming pod guest, Rick Faulk — this time alongside Roy Belak, the CEO of Nexera, a startup recently scooped up by Locus. Rebecca, meanwhile, chatted with George Morton, founder of solar-powered autonomous boat startup, Open Oceans.
All that, plus a new Startup Spotlight with an industrial window washing robot. What are you waiting for? Let's get Automated.
A year before Array made its debut on the floor at Modex, viewing the robot was an invite-only affair.
“Back in 2025, Array was a prototype,” says Rick Faulk, “a good prototype, but we knew we had to enhance that. Frankly, the question we got over and over and over as we did demos behind the wall at that show was, ‘Will it pick my thing?’ That was the first question that we got. Folks loved the technology, but literally the first question out of their mouth was, ‘Will it pick it?’ ”
The mobile picking system, which represents a punctuated evolution for Locus’ AMR offerings, was the subject of select, behind-the-curtain showcases at major industry tradeshows. The automation firm previewed the robot for select partners, potential customers, and members of the press.
“We answered them truthfully, saying, ‘Yeah, we have a little more, a little more work to do on the grasping side,’ ” Locus’ CEO adds, “and to ‘stay tuned.’ ”
The version of Array shown off in previews utilized a fairly traditional suction-based end effector for picking and placing inventory. Ditto for the multiple models that appeared on the Modex show floor a couple of months back. All the while, Locus had what it believed was the ideal solution waiting in the wings.
Ideally, the Arrays cruising the floor of Atlanta’s Georgia World Congress Center would have showcased Nexera’s NeuraGrasp technology. But hey, turns out there’s a lot of paperwork involved in buying a startup — particularly in another country (even if that country happens to be our neighbor to the north), and timelines don’t always line up perfectly. The good news, however, is you suddenly have a new opportunity to talk to the press about your cool new robot.
“My opinion is that there's too many opinions about robots out there,” Kate Darling notes. Such a sentiment can go either way, really, roughly one-third of the way into an hour-long podcast interview that is, for all intents and purposes, about robots.
There are, objectively, a lot of opinions about robots out there. I suspect I encounter more than most do as part of my day-to-day operations. Statistically speaking, some are better than others, though who serves as the ultimate arbiter is another question altogether.
We can, to a certain extent, analyze these opinions based on available data. In cases where opinions take the form of predictions, however, time will be the ultimate arbiter. In the meantime, I recommend taking up hobby as diametrically opposed to your day job as you’re comfortable with.
“I didn't want to add more opinions,” Darling adds, laying out the foundations of the RAI Institute’s Robotics, Ethics & Society research team. “So I specifically hired people who are trained in different methods of data collection. Outward-facing research and mission is to collect data to better inform decision-making in the design or integration, regulation, or standard settings of robotics.”
Prior to leading a team at RAI, Darling spent 14 years as a research scientist at MIT’s Media Lab, with a focus on human-robot interaction. Her HRI journey began a few years prior to that, in 2008, when a robot dinosaur arrived on her doorstep from Hong Kong. “Pleo is the whole reason that I got interested in robotics,” she explains.
But Daniela Rus says the future of robotics will not be defined by viral humanoid robot demos alone. The real challenge is building robots that can understand the physical world, make safe decisions in real time, and work reliably outside controlled lab environments. Watch on YouTube>
The Cartwheel Robotics story ended on a down note roughly three months back. Founder Scott LaValley landed on his feet, of course, ending up on a growing team over at Google DeepMind. When he announced that his startup’s days were drawing to a close, the former Disney Imagineer wrote cryptically, “To all the founders out there, a hard-earned lesson (please listen): No money is better than the wrong money.” Now we’re getting more insight, as the startup grapples with an “involuntary” Chapter 7 proceedings.
LaValley has opened the books on his own hard-learned lessons in a bid to help other early-stage founders avoid similar pitfalls. “These filings are not the whole story,” he writes. "They are the opening public record in a case that will continue to develop through the court process over many months. Some filings include allegations, disputed characterizations, and procedural positions by parties in interest. The archive is not the final word — it is the public record to date.”
Check out the University of Michigan Winter 2026 Capstone projects. Per the school, "Undergraduate students work with an external group, such as Subaru, to study, design, build, and iterate on a solution to an open-ended robotics problem." It’s a diverse bunch. You’ve got a mobile robot that can sense the source and direction of smells, a system for monitoring cervical dilation during labor, control for swarms of magnet robots, and a tool that measures the torque of jerboa tails. If you weren’t already familiar with the waifish desert rodents, consider this an excuse to go down an internet jerboa hole this afternoon.
Genesis AI has kickstarted the press release flywheel. Seems every week it’s got something new to blog about. Today, it’s the launch of Genesis World 1.0, which builds on the simulation platform the physical AI firm unveiled a while back, prior to launching in its present startup form. New Github entries reveal a very lifelike sim platform (that's a sim image above, BTW), which Genesis AI says should condense “a full week of nonstop robot testing [… into] about 30 minutes in simulation, with consistent and repeatable results." The startup adds, "Simulation is not merely a data generator for robotics. It is a fundamental infrastructure layer. Once simulation is embedded into the development process, progress can move at the speed of compute, rather than being limited by real-world wall-clock time.
This week, Rebecca chats with Online Oceans’ George Morton about the U.K. startup’s autonomous, solar-powered data-collecting boat, Scout. The cofounder and CEO compares the system to a Ford F-150, which makes things much easier for my American brain to comprehend. Also some interesting founding tidbits in here, including the org’s connection to the Microtransat Challenge, a DARPA-style challenge involving autonomous boats.
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.
