Cambrian Robotics demonstrated vision-guided automotive assembly at Automate. Source: Oliver Mitchell
May and June were big months for robotics in the U.S. As flowers bloomed outside, actuators rose in the convention halls of Boston, Detroit, Houston, and Atlanta. While there have been many good articles on each of these events, with numerous postings of new humanoids and cobots in action, I would like to share some observations as an insider with nearly a decade of attendance at two very different shows: Automate and Xponential.
On a personal note, since starting this article, uncrewed systems that were demonstrated on the floors of these shows weeks earlier have now been deployed globally on battlefields and on parade grounds in Washington, D.C.
Automate grows up
In its heyday, Detroit was the fourth-largest city in the U.S., manufacturing 75% of the country’s automobiles. Today, its rank has fallen to 27, claiming one of the lowest per capita incomes in the country and producing fewer than 20% of our cars.
Last month, as I walked the halls of Automate with robot arms in trade show booths vibrating, spinning, and compressing their loud motors, I left with the feeling that the Motor City could experience an urban renaissance led by A3, or the Association for Advancing Automation.
“Our data across a 30-year period tells us that robots are helping to save and create jobs,” said Jeff Burnstein, A3’s president, when asked about the impact of robots on jobs. “When robot sales increase, employment also rises, and vice versa. The real risk to jobs is when companies can’t compete, as we see from the empty factories that are so visible right here in Detroit.”
“Advances in technology mean jobs will be different. But that’s always been the case: 30 years ago, there was no such thing as a digital marketing manager or search engine optimization specialist,” he added. “Robots are tools to help companies improve productivity, increase quality, speed time to market, and ultimately win more business and hire more people. The jobs of the future will be better, safer, and higher paying.”
Automate has grown up. What started as a dog leg of ProMat is now its own annual show. A3’s event has already outgrown Detroit’s Huntington Place Convention Center, with over 45,000 attendees crowding the space. Note that Automate will return to Chicago in 2026, taking over McCormick Place.
Now accepting session submissions!
Interest in Cambrian explodes
This year, not a single corner of the massive trade show floor remained open. For four days, robots reigned supreme over the city, with every manufacturer making the pilgrimage to see the latest technology that might advance their domestic workflows amid a trade war.
I witnessed this firsthand while working the booth of portfolio company, Cambrian Robotics, with a steady stream of automotive engineers arriving on tour buses to catch a glimpse of AI-infused, 3D computer vision technology.
Cambrian’s suite of workflow applications, using a wide range of OEM robot arms, garnered attention from prospective users. Automotive executives saw the speed and ease of robotic brake assembly, metal-hinge bin picking, and picking and placing of transparent objects.
The biggest draw, possibly even surpassing Agility Robotics’ humanoids, was Cambrian’s two-arm UR wire harness and insertion demo (see below). This example illustrates the current capabilities of large language models (LLMs) in the field of robotics.
This simulation feasibility study trains robots in even more scenarios than they will encounter in the real world, thus enabling the quick deployment of deterministic, mission-driven systems that complete the job, regardless of manipulation mishaps. This new wave of AI-driven products was not limited to Cambrian or other computer vision offerings and was a general theme of this year’s shows.
Startups advance with AI
I’ve judged the Automate startup competition almost consistently since the show’s inception. This year may have been the most impressive group of startups, including one backed by robotic luminary Daniel Theobald.
The cohort included numerous companies utilizing foundational AI models to push the envelope of robotics. Two standouts were Kinisi (the winner) and Nexus Intelligence (my runner up).
Kinisi has developed a wheeled, two-armed robot with a small head appendage that hinges up and down. Nine months earlier, the founder Brennand Pierce, posted on LinkedIn about his newest product, which uses OpenAI to learn new behaviors.
“The key takeaway from the video is that all the robot’s behaviors are generated by the LLM,” he asserted. “It only knows a few basic functions: open/close gripper, move arm, and process input from the vision pipeline. From these, it can interpret my voice commands and combine them into more complex tasks. In this example, it passes me a ‘cold drink’ by recognizing a can of Coke in front of it.”
Nexus, a deserving honorable mention, developed a generative AI platform to connect industrial automation control systems, such as programmable logic controllers (PLCs), to co-pilot programming platforms, thereby speeding up integrations.
I’m sobered by Rodney Brooks’ view on LLMs, expressed last year in TechCrunch: “People say, ‘Oh, the large language models are gonna make robots be able to do things they couldn’t do.’ That’s not where the problem is. The problem with being able to do stuff is about control theory and all sorts of other hardcore math optimization.”
“It’s not useful in the warehouse to tell an individual robot to go out and get one thing for one order, but it may be useful for elder care in homes for people to be able to say things to the robots,” he added.
Insights from Xponential
Moving from Detroit, I landed in Houston (the current fourth-largest American city) for AUVSI’s Xponential to host a panel on defense technology. Similar to Automate, this show has exploded, with new commercial exhibits and applications filling the entire George R. Brown Convention Center.
Since attending this show nearly a decade ago, I’ve noticed a shift in attendee demographics, with a growing number of commercial buyers and partners now outnumbering uniformed military personnel and public service officers. In fact, it is now promoted as “the largest trade show and conference for both commercial and defense UAVs, autonomous vehicles, and robotics.”
While the space was filled with eVTOLs, winged UAVs, terrestrial robots, and numerous new marine innovations, the busiest part of the show was still the Defense Tech theater.
Since the launch of the Defense Innovation Unit by former Secretary of Defense Ash Carter, the relationship between the startup community and the military has evolved, empowering warfighters with less expensive mobile, uncrewed systems, such as drones and packbots, outside the traditional supply chains of Defense Primes.
Ironically, global conflicts have become advertisements for the upstarts in this fast-moving environment, from Ukrainian companies whose drones on display in Houston destroyed Russian bombers weeks later to Israeli cyber technologies that are attracting $100 million seed round funding from top-tier firms like Lux Capital and Sequoia.
I asked Dan Madden of Squadra VEntures and Jacqueline Blackburn of Decisive Point about their views on how a generation raised on ChatGPT will change the internal operations in the Defense Department. They retorted that this is what their funds are banking on with their portfolios within the robotics, drone, and intelligence arenas.
Earlier this month, defense tech startup Anduril raised a $2.5 billion Series G round, doubling its valuation to $30.5 billion. The round was reportedly 8x oversubscribed, with many investors being turned away. This raise highlights the potential for the defense tech market, particularly in the realm of autonomy.
“So we’re building a lot of different things, but I’d say the thing that is going to dominate capacity is going to be autonomous fighter jets,” explained Palmer Luckey, founder and CEO of Anduril, to The Wall Street Journal. “We are building an autonomous fighter jet for the U.S. Air Force called Fury for the CCA or Collaborative Combat Aircraft program.”
“This is a really big win that Anduril had last year, where we were competing against a number of different companies, including Lockheed Martin, Northrop Grumman, and Boeing,” he noted. “We beat the big guys and managed to convince the Air Force that we were the ones who are going to be able to build the best autonomous fighter jet and that we were going to be able to build them in a scale of thousands of units on a timeline that is relevant to a potential fight in the Pacific with China.”
Autonomous fighter jets are like drones on steroids, led by a startup that didn’t exist eight years ago. I’m looking forward to attending next year’s Xponential, which will be held in Detroit.
Oliver Mitchell hosted a Defense Tech panel with Dan Madden of Squadra Ventures and Jacqueline Blackburn of Decisive Point. Source: Oliver Mitchell
Back to Boston for Robotics Tech Week
I ended my robotics trade show season in Boston, attending Robotics Invest. While Cybernetix Ventures was not the only sponsoring entity, it is its brainchild.
Fady Saad, general partner at Cybernetix, curated a panelists and guests, with three ff Venture Capital portfolio companies in attendance this year: Plus One Robotics, Cambrian Robotics, and Burro.
I was privileged to launch my book tour at MassRobotics with 100 people in attendance, sponsored by Lowenstein Sandler. In many ways, the interviews in my Startup Field Guide parallel the growth of the industry, from Mick Mountz’s revolution in e-commerce fulfillment to Nic Radford’s latest humanoid invention.
Who knows? Maybe the next unicorn founder walked out with a signed copy of A Startup Field Guide in the Age of Robots and AI to begin their own robo-adventure.
Oliver Mitchell’s book launch at MassRobotics with Marita McGinn during Boston’s Robotics Tech Week.
The post VC reflects on robotics trade show season appeared first on The Robot Report.