Universal Robots Redesigns Its Cobot Platform for the Physical AI Era

Universal Robots unveiled Gen 7 on Wednesday in Denmark, a ground-up redesign of its collaborative robot platform built around a new controller, teach pendant, and tool flange meant to let AI models plug directly into a cobot arm rather than bolt onto one. The launch introduces three new arms, the UR10g-1750, UR17g-1300, and UR18g-950, alongside a CB7 Core controller and TP7 Core pendant, and it marks the first time Teradyne's cobot unit has redesigned every layer of its hardware and software stack in a single release rather than iterating a single component.

For a procurement team that has spent the last two years fielding vision add-ons, force sensors, and third-party AI kits as bolt-on afterthoughts to a decade-old cobot architecture, Gen 7 is Universal Robots' argument that the bolt-on era is over. The company says the CB7 Core packs 40 percent more compute in a 30 percent smaller footprint than its predecessor, with multi-network communication built in, while the TP7 Core pendant is 20 percent lighter with a full HD display and configurable buttons. The real change sits underneath: an AI-ready tool flange that exposes real-time vision and sensing data directly to the controller, plus built-in force-torque sensing and impedance control, so a robot can feel resistance and adjust mid-motion instead of running a fixed trajectory and stopping on a fault.

"The robots of the AI era must be advanced, not complicated," said Jean-Pierre Hathout, president of Teradyne Robotics Group, in the announcement. "Gen 7 is an end-to-end redesign." The line captures the commercial bet Universal Robots is making: that the next wave of cobot buyers will not tolerate stitching together a vision system from one vendor, a force sensor from another, and a software layer from a third just to get a robot to react to an unstructured environment. Gen 7 folds that integration into the base platform and prices it as one purchase decision instead of three.

The Software Layer Buyers Actually Asked For

The most consequential change in Gen 7 is not a spec sheet number. It is PolyScope X, the new operating system running on the CB7 Core, which adds native ROS 2 communication and open application programming interfaces (APIs) on top of the existing ISO 10218-1 safety certification that industrial buyers require for any collaborative robot sharing floor space with people. ROS 2 support matters commercially because it removes a translation layer that integrators have had to build by hand for years: a warehouse or electronics assembler running a ROS 2-based perception stack can now talk to the arm's motion controller without a custom bridge, which shortens the deployment timeline that has historically made cobot pilots take months rather than weeks.

That timeline compression is the actual product Universal Robots is selling. Cobots differ from traditional caged industrial arms in that they are marketed on ease of deployment as much as raw throughput, and Universal Robots built its position over two decades and more than 100,000 installations worldwide on that premise. Gen 7's bet is that "ease of deployment" now has to include AI integration, because the buyers evaluating cobots in 2026 are increasingly asking not just "how fast does it move" but "can it use a camera and a language model to figure out what to pick up." A tool flange that streams vision and force data straight into the controller, rather than through a separate industrial PC bolted to the arm's base, answers that question at the hardware layer instead of leaving it to a systems integrator to solve on-site.

Reading the Timing Against the Parent Company's Numbers

Universal Robots operates as the cobot division of Teradyne, the semiconductor test-equipment maker that acquired the Danish company in 2015, and Teradyne's robotics segment has been the more volatile half of that portfolio during the AI hardware buildout of the past two years. A redesign that touches every layer of the product at once is a higher-risk release than an incremental refresh: it means the entire installed base of integrators, systems houses, and training programs built around the prior generation's controller architecture now has to requalify against PolyScope X. Universal Robots is clearly betting that the commercial upside of a unified AI-ready platform outweighs that transition cost, and the decision to launch three new arm variants simultaneously rather than staggering the rollout suggests the company wants to close the gap with AI-native entrants quickly rather than let Gen 7 trickle into the lineup over a year.

The three new arms span a deliberate range: the UR18g-950 favors reach and payload for heavier material handling, the UR10g-1750 keeps the long-reach profile that made the UR10 a warehouse and logistics staple, and the UR17g-1300 sits in between for general assembly work. None of the three specification sheets differs dramatically from the outgoing g-Series lineup on raw payload or reach. The differentiation is entirely in what the arm can perceive and how quickly it can be reprogrammed, which is a deliberate signal that Universal Robots does not see its next competitive battle being fought on payload class. It sees it being fought on how fast a customer can go from unboxing an arm to running an AI-guided pick-and-place task without hiring a robotics integrator to write custom middleware.

What Buyers Still Do Not Know

Universal Robots has not disclosed pricing or a firm shipping date for any of the three Gen 7 arms, which leaves procurement teams currently budgeting 2027 automation lines without the one number they need most. Collaborative robot pricing has historically tracked close to the outgoing generation at launch, with premiums fading as volume ramps, so buyers evaluating Gen 7 against rival platforms from Yaskawa, ABB, or Fanuc should expect list prices in line with the current g-Series family rather than a steep AI premium, though that remains Universal Robots' assumption to confirm, not a buyer's guarantee. The bigger open question is backward compatibility: whether existing UR fleets running the older PolyScope software can be field-upgraded to any part of the CB7 Core stack, or whether Gen 7 is a clean-sheet product line that leaves five-year-old installed cobots on a separate support track. Universal Robots has not addressed that question in the launch materials, and it is the detail that will determine whether Gen 7 is an upgrade path for the existing customer base or a parallel product line aimed at new buyers.

Universal Robots has built three years of AI partnership announcements, from vision-model integrations to compute partnerships, that never quite closed the gap between a demo booth and a certified factory-floor product. Gen 7 is the first release that puts those capabilities into hardware customers can actually order, which makes it a materially different milestone than another software update layered onto five-year-old arms.

Where This Leaves the Rest of the Cobot Field

Every major collaborative robot maker has spent 2026 making some version of the same argument Universal Robots is now making in hardware. Fanuc, Kawasaki, and ABB have each paired their arms with Nvidia's Isaac stack for perception and simulation, while Yaskawa has leaned on force-sensing upgrades to its Motoman line to compete on the same "feel, don't just follow a path" proposition that Universal Robots is now building into Gen 7's tool flange as a standard feature rather than an option. What separates Universal Robots' move is scope: rather than adding one AI-adjacent feature to an existing arm, it redesigned the controller, the pendant, and the mechanical interface simultaneously, which is a more expensive bet for the company but a simpler one for a buyer trying to evaluate a single coherent platform instead of a checklist of add-on modules from competing suppliers.

That scope also raises the qualification bar for the rest of the industry. Once one of the two or three largest cobot vendors by installed base ships native ROS 2 support and an AI-ready sensing interface as the default configuration rather than a premium tier, competitors that keep those capabilities as paid add-ons will have a harder time justifying the price gap to buyers who can now get them bundled. Universal Robots has effectively reset what "current generation" means for cobot procurement, and the vendors still selling architectures built around bolt-on vision kits and aftermarket force sensors now have to explain why their roadmap has not caught up.

What a Procurement Team Should Actually Check

For a plant manager or automation buyer evaluating Gen 7 against an existing fleet or a competing quote, three questions matter more than the headline specifications. First, whether a systems integrator already certified on the current UR software stack needs retraining on PolyScope X, since that retraining cost sits outside the sticker price and can erase a chunk of any efficiency gain in year one. Second, whether the AI-ready tool flange requires proprietary Universal Robots sensing hardware or accepts third-party vision and force sensors through an open interface, which determines whether a buyer is locked into a single vendor's sensor ecosystem or free to shop the broader market. Third, whether the ISO 10218-1 safety certification that PolyScope X carries forward has been re-validated for the specific combination of a Gen 7 arm and a customer's existing safety-rated monitored stop configuration, since collaborative robot deployments near human workers cannot skip that step regardless of how capable the new controller is on paper.

None of those three questions has a public answer yet, and Universal Robots will need to close that gap before Gen 7 orders convert from interest to signed purchase agreements. A platform redesign this comprehensive tends to sell on the promise of a shorter integration timeline, and the buyers who move first will be the ones who verify that promise against their own safety and retraining costs rather than the announcement's compute and weight figures alone.

This analysis draws on Universal Robots' product announcement and public statements from Teradyne Robotics Group. It is for general information purposes only and does not constitute investment, financial, or legal advice.

Hero image credit: Universal Robots / Teradyne Robotics Group.