What Happens When Robots Die? The Case for End-of-Life Standards in Robotics
Why robot shutdowns should be planned like product launches.
Aaron’s Thoughts On The Week
“The measure of a society is how it treats its most vulnerable... including the ones it built.” — Adapted from Hubert H. Humphrey
Late last year, startup Embodied announced it was shutting down and taking its flagship product, the Moxie robot, with it.
Moxie, an $800 AI-powered companion designed to support the emotional development of children ages 5 to 10, quickly became a beloved presence in many homes. But with Embodied’s cloud services going offline, the robots would soon be bricked, unable to perform even basic functions.
The backlash was immediate and emotional. In response, Embodied’s CEO Paolo Pirjanian announced via LinkedIn that a group of former engineers was working on an open-source recovery effort called OpenMoxie. This would let users run a local server on their computers, keeping basic features alive without relying on the company’s now-defunct infrastructure. An OTA (over-the-air) update had been released to prepare Moxie units for this possible future.
Now, OpenMoxie is officially available. The project enables Moxie robots to function without Embodied’s servers, which shut down on January 30, 2025. However, there’s a catch: users must have installed the critical OTA update (version 24.10.803) before the shutdown. This update prepares Moxie to connect with a local OpenMoxie server, allowing continued access to features like Daily Missions and Reading activities.
While this was a hopeful outcome, it remains community-supported and unsupported by Embodied. Many users may have missed the update window or may be unable to complete the setup, highlighting how fragile robot ownership can be when vendors fail.
Unfortunately, Moxie is not an isolated case.
Aldebaran Robotics, once a global leader in social robotics, suffered a similarly quiet collapse this week. It’s NAO and Pepper robots that are used in classrooms and customer service around the world. It is too early to determine what support customers will receive after the bankruptcy, but it does not look promising currently. There is a possibility that another company will acquire Aldebaran's assets and continue to support Pepper and NAO customers, provided the new owner is willing to do so. The last Pepper was manufactured over 4 years ago.
Moxie, Pepper, and NAO will join the likes of Jibo, Anki, Kuri, and the growing list of unsupported robots that are out there in the world. As these cases accumulate, it is becoming clear that the robotics industry lacks responsible end-of-life practices.
Why End-of-Life Standards Matter
As robots move beyond novelty into real roles in homes, classrooms, hospitals, and warehouses, sudden shutdowns are no longer just disappointing; they’re disruptive and expensive.
To address this growing gap in accountability, the robotics industry must adopt end-of-life standards. These should include:
Escrow and Transparency from Day One
Vendors should clearly communicate the operational dependencies of their robots, especially concerning cloud services and potential offline functionalities.
The sudden shutdown of Embodied left Moxie robot owners in a lurch. The robot’s core functionalities, Daily Missions, reading exercises, and interactive conversations, were all dependent on Embodied’s cloud infrastructure. Without it, Moxie could do little more than blink. While the company eventually announced OpenMoxie, a community-driven project to give the robot a second life, this came only after significant user backlash. Worse still, Moxie owners had to manually download a critical over-the-air (OTA) update before the company’s servers shut down—something many users missed.
This reactive response exposed a fundamental failure in planning. Had Embodied employed a software escrow model, the situation could have been very different. In enterprise IT, software escrow ensures that a trusted third party securely holds a copy of the software source code, system documentation, and deployment tools. If the vendor dissolves or fails to meet service obligations, customers can legally access the escrowed materials to maintain or replicate the product independently.
If a software escrow had been in place for Moxie, robot owners, especially institutions like schools or clinics, could have continued to operate the system without relying on the company’s goodwill or volunteers scrambling to patch together open-source alternatives. Instead, families who invested in Moxie’s promise were forced into a race against a shutdown clock, with little technical support and even less clarity.
For a robot sold as a long-term emotional support companion for children, that’s not just poor planning, it’s a breach of user trust. In the future, embedding software escrow provisions into robotics business models and procurement contracts should be a best practice, especially for any robot that depends on external servers to function.
Modular, Open Design: Building Robots That Can Outlive Their Makers
One of the biggest lessons from the collapse of companies like Anki and Embodied is that when a robot is a tightly closed system, its fate is entirely tied to the company that made it. But when it's modular and built on open standards, it gains resilience, and its users gain freedom.
A modular architecture means that individual components—such as sensors, actuators, processors, or software modules—can be swapped out, repaired, or upgraded independently. And when those components adhere to widely adopted standards, developers and hobbyists alike can plug in alternatives or expand functionality without reinventing the wheel.
That’s where open platforms like ROS (Robot Operating System) and URDF (Unified Robot Description Format) come in. ROS, an open-source framework, has become the de facto backbone for research and prototyping across the robotics world. URDF, the standardized robot modeling language, allows users to describe physical characteristics, such as joints and links, in a way that can be easily reused across simulators and hardware.
These tools aren’t just popular, they’re sustainable. Entire communities have sprung up around them. Resources like AutomaticAddison.com and hundreds of GitHub repositories offer templates, examples, and tools that dramatically reduce development time and increase longevity.
Robots like SMARTmBOT and ROMR exemplify this approach. Built on open-source frameworks, these platforms are designed for adaptability. Want to add an arm? Swap in a different vision system? Update the software stack entirely? You can—and others probably already have.
Contrast that with the fate of robots like Cozmo or Vector from Anki. When the cloud services powering their personalities were shut off, they lost core functionality. The only way forward was through complex reverse engineering and scattered community efforts, many of which struggled due to a lack of documentation and proprietary constraints.
By designing robots with modularity and openness in mind, we create machines that can grow with their users, evolve, and continue to function long after the original vendor has exited the scene.
Ethical Exit Protocols: Planning for a Robot’s Retirement
Robots, like any product, have a lifecycle. They’re imagined, designed, built, deployed, and eventually, they become obsolete. But in robotics, obsolescence isn’t just a technical matter. It’s ethical.
What happens when a robot that helps care for an aging parent stops receiving software updates? Or when a classroom teaching assistant is suddenly unsupported, leaving students and educators scrambling? The industry’s answer, far too often, has been: nothing.
We don’t let critical medical devices disappear without backup plans. So why do we accept it with robots that manage safety, education, or emotional well-being?
That’s where the idea of ethical exit protocols comes in. It’s about embedding responsibility into the lifecycle of robotic systems from the start. That means thinking not just about how to launch a robot, but how to sunset it safely, transparently, and humanely.
Some groundwork already exists. The ISO 13482 standard, for example, focuses on safety requirements for personal care robots, including those that assist with mobility, communication, and interaction. While largely framed around physical safety, the standard implicitly points to a broader principle. When a robot is used in intimate or support-based roles, the stakes for failure or abandonment are higher.
Similarly, UL 3300 establishes safety benchmarks for consumer, education, and service robots. It pushes manufacturers to consider not just immediate safety, but long-term usability, data management, and support pathways throughout a robot’s life.
But standards alone aren’t enough. They need teeth and adoption. What’s missing is a clear industry-wide commitment to responsible retirement planning for robotic systems. That could include things like:
Minimum timelines for support after product discontinuation
Requirements to publish documentation and APIs when the service ends
Legal obligations to notify users and offer migration tools
Inclusion of third-party support or transfer options as part of product design
Companies building robots for children, patients, or frontline workers have a duty of care to ensure the safety of these individuals, both physically and mentally. It's not enough to innovate; they must also be accountable when their innovation comes to an end.
Too often, the people most affected by a robot’s demise are those with the least ability to replace it quickly: small schools, underfunded clinics, or families on fixed incomes. Without ethical exit strategies, these users are left unsupported and often unfairly punished for being early adopters.
The way forward is clear: build ethical sunsetting into the foundation of robotics business models. Not every robot can last forever, but every robot should leave behind a plan.
Third-Party Support Ecosystems: Keeping Robots Alive After the Warranty Ends
Robots, like any technology, need maintenance. But too often, repair and support are locked behind company walls. When a robotics company shuts down or abandons a product, users are left with little more than expensive paperweights—unless there’s a broader ecosystem to step in.
That’s where the Right to Repair movement comes in. This growing global effort advocates for laws that require manufacturers to provide the necessary tools, parts, and information to repair their products. States like New York and California have passed such laws, and Europe is moving even faster.
Robotics should follow suit. Devices that blend proprietary hardware with cloud-based AI can become inoperable without access to diagnostics or updates. We’ve seen this with robots like Cozmo, Jibo, and Moxie, where only after public backlash or volunteer action did partial solutions emerge.
Initiatives like the Open Source Robotics Alliance (OSRA) are laying the groundwork for a more effective model, supporting open-source tools that anyone can build upon. Companies should follow that lead by publishing service manuals, APIs, and software access to empower repair shops, educators, and hobbyists.
Because a robot’s life shouldn’t end when its company does.
We Need to Plan for the End
Not every robot needs to last forever. But every robot user, whether a parent, teacher, researcher, or business, deserves to know what to expect if it doesn’t.
Moxie was marketed as a child’s emotional companion, a helper kids could rely on daily. When Embodied abruptly shut down, families were left scrambling, many of whom had no time to install a critical update. A promised presence became a silent, blinking shell.
And Moxie is not alone. We've seen countless robots, some beloved, some mission-critical, vanish without notice. With little to no support, users are left with broken promises, expensive hardware, and no roadmap forward.
These robots aren’t just tools. They become part of workflows, lesson plans, and lives. When they disappear without warning, the impact is real—financially, operationally, and emotionally.
The robotics industry needs to mature. In a world where robots are entering classrooms, hospitals, homes, and public spaces, responsible design must include responsible endings.
That means clear communication, open documentation, escrow options, modular design, and ethical sunsetting standards. The pieces are all there, it’s time to put them in place.
Because when a robot dies, the damage shouldn’t ripple outward to the people who depended on it.
The future of robotics isn’t just about how we build them. It’s about how we let them go.
Robot News Of The Week
Congressional Robotics Caucus relaunches to help U.S. industry
The Congressional Robotics Caucus has officially relaunched, with bipartisan leadership from Reps. Jim McGovern (D-MA), Bob Latta (R-OH), Haley Stevens (D-MI), and Jay Obernolte (R-CA). The group aims to educate lawmakers and shape policies around robotics, automation, workforce development, and U.S. competitiveness.
After being largely inactive since 2019, the caucus returns at a time when the U.S. lags in global robot density and faces increasing pressure from countries like China, which is investing heavily in robotics. The caucus will focus on issues such as ethical deployment, national strategy, education, and public-private R&D partnerships.
Supporters hope this renewed effort will help the U.S. maintain global leadership in robotics innovation.
HEBI Robotics has received a $250,000 U.S. Army SBIR grant to develop modular robotic components certified for use in hazardous environments with flammable vapors or liquids (NEC Class I Division 1). Most robots aren't built for these explosive settings, making remote operation difficult and costly.
HEBI’s project aims to simplify and reduce the cost of building certified robotic systems for hazardous industrial and military tasks, such as fuel tank inspections and handling volatile materials. CEO Bob Raida calls this a significant step toward enabling professionals to build rugged, high-performance robots for challenging conditions.

UK medical robotics company aims for $4bn sale
UK-based CMR Surgical is exploring a potential sale, seeking a valuation of up to $4 billion as it prepares to expand into the U.S. — the world’s largest surgery market.
Founded in 2014, CMR competes with firms like Intuitive Surgical and has raised $200M this year to support the U.S. launch of its portable surgical robot, Versius, which has already been used in over 30,000 procedures globally.
The sale would likely attract major strategic buyers, though no final decision has been made. The move highlights ongoing challenges in scaling UK life sciences firms without foreign acquisition.
Comau’s Robotic Systems Take on 3D Printing Applications
Comau, the Italian robotics firm owned by Stellantis and One Equity Partners, is rapidly expanding its presence in additive manufacturing (AM) through diverse global partnerships. From helping CEAD 3D print boat hulls to collaborating with Prima Additive and Titomic on metal and coating applications, Comau’s robotic arms are enabling innovations across the maritime, automotive, and construction sectors.
These moves reflect Comau’s strategy to meet growing demand for advanced, localized manufacturing solutions, especially relevant amid global supply chain disruptions. Its focus on automotive applications, in particular, could play a key role in reshoring production and supporting Western competitiveness in EV design and manufacturing.
Robot Research In The News
Give Bots a #@%! Chance: The Case for Cursing Robots
Today’s robots are designed to be polite, obedient, and non-threatening, but this may reinforce outdated norms that associate helpfulness with subservience. At Oregon State University, researchers decided to challenge these assumptions by experimenting with robots that swear when they make mistakes.
They tested public reactions to robot errors—like dropping objects—with three types of responses: silence, polite speech, or profanity. Surprisingly, the results showed that cursing didn’t harm public perception and, in fact, increased humor and social closeness among university students. The general public was slightly more reserved, but still rated swearing robots similarly to non-swearing ones on most measures.
A follow-up in-person study confirmed the findings: people appreciate it when robots acknowledge their mistakes, regardless of whether they use profanity. However, context matters; for instance, swearing around children or in formal settings may not be appropriate.
Despite strong data and positive reviewer feedback, the research faces resistance in publication. But the team hopes designers will consider more expressive, even foul-mouthed robots, used thoughtfully, to create more relatable interactions.
Robot Workforce Story Of The Week
New Robotics and Mechatronics Degree Debuts at Randolph College
Randolph College in Lynchburg has introduced a four-year degree in robotics and mechatronics, combining traditional engineering with advanced technologies such as 3D printing and robotic arms. The program, which debuted last fall, provides students with immediate hands-on lab experience and emphasizes both technical skills and liberal arts foundations, developing critical thinking and communication skills alongside engineering.
Professor Peter Sheldon highlights the value of Randolph’s smaller size: “At big schools, undergrads rarely touch the equipment. Here, they dive in right away.” Graduates will be prepared for careers in fields from biotech to national security, with a focus on innovation and adaptability.
Robot Video Of The Week
Scientists created a soft robot that moves like snakes or worms by bending its body and using friction to crawl. The robot is composed of soft, inflatable parts and a specialized outer skin that enhances its grip on the ground. It can move forward by pumping air into different parts of its body and turn by changing how it moves on each side. The robot was tested in a challenging area with rough surfaces and obstacles. Using sensors and human control, it avoided crashes. This type of robot could be helpful in search-and-rescue missions or exploring tight, confined spaces.
Upcoming Robot Events
June 9-13 London Tech Week
June 17-18 MTC Robotics & Automation (Coventry, UK)
June 30-July 2 International Conference on Ubiquitous Robots (College Station, TX)
Aug. 17-21 Intl. Conference on Automation Science & Engineering (Anaheim, CA)
Aug. 25-29 IEEE RO-MAN (Eindhoven, Netherlands)
Sept. 3-5 ARM Institute Member Meetings (Pittsburgh, PA)
Sept. 15-17 ROSCon UK (Edinburgh)
Sept. 23 Humanoid Robot Forum (Seattle, WA)
Sept. 27-30 IEEE Conference on Robot Learning (Seoul, KR)
Sept. 30-Oct. 2 IEEE International Conference on Humanoid Robots (Seoul, KR)
Oct. 6-10 Intl. Conference on Advanced Manufacturing (Las Vegas, NV)
Oct. 15-16 RoboBusiness (Santa Clara, CA)
Oct. 19-25 IEEE IROS (Hangzhou, China)
Oct. 27-29 ROSCon (Singapore)
Nov. 3-5 Intl. Robot Safety Conference (Houston, TX)
Dec. 11-12 Humanoid Summit (Silicon Valley TBA)