It’s Not the Robot, It’s the Application: A New Era in Standards

Language matters: ISO and ANSI clarify that collaboration is about context, not a robot’s marketing label.

“The beginning of wisdom is to call things by their proper name.” – Confucius

For over a decade, “collaborative robot” – often affectionately shortened to cobot – has been a buzzword in automation. It described a new breed of robots designed to work safely alongside humans, free from the confines of traditional industrial robots. The term entered wide use in the late 20th and early 21st centuries as companies explored increased human-robot interaction. However, the latest standards update has given this popular term a reality check (and a formal send-off). In 2025, international and U.S. standards bodies decided that collaboration is something you do, not something a robot inherently is.

The revised ISO 10218 (2025) – the flagship international robot safety standard – pointedly omits any definition of "collaborative robot." In a published FAQ, the standards committee makes it clear: “The terms ‘collaborative robot’ and ‘collaborative operation’ will not be found in the revised ISO 10218. ‘Collaborative application’ is used instead, as only the actual use of the robot can be designed, tested, and confirmed as a collaborative application.” In other words, you won’t find the word cobot in the new ISO rulebook at all. The American National Standards (ANSI/A3) followed suit. The updated ANSI/RIA R15.06-2025 (the U.S. industrial robot safety standard adapted from ISO 10218-1/-2:2025) likewise eschews the term “collaborative robot,” replacing it with the more precise concept of a collaborative application.

This terminology change might sound like splitting semantic hairs, but it carries significant weight. It reflects a maturation in how experts think about safe human-robot interaction. After all, simply calling a machine a collaborative robot doesn’t magically make it safe in all scenarios, and the standards developers finally had enough of the confusion.

Why “Collaborative Robot” Got the Boot

Standards committees are not known for making frivolous edits. So why purge “collaborative robot” from the vocabulary? The short answer: clarity and safety. The old term had become misleading, even marketing-driven, and experts worried it gave end-users a false sense of security. As one safety professional put it, “The ‘collaborative robot’ term has mistakenly become synonymous with robots utilizing power and force limiting technology.” In other words, people started to equate cobot purely with one built-in safety feature (power-and-force limiting, or PFL), ignoring the broader context of how the robot is deployed.

In reality, there is no special breed of robot that is automatically safe just by virtue of its label. “In actuality, there is no such thing as a ‘cobot,’ rather there are robots that utilize collaborative technologies,” explains Todd Dickey, chair of the R15.06 drafting committee. Those collaborative technologies defined in the standards include:

• Hand-guiding controls – allowing an operator to physically lead the robot’s motions;

• Speed and separation monitoring (SSM) – using sensors to keep a safe distance or slow the robot when humans approach;

• Power and force limiting (PFL) – built-in limits so the robot can safely bump into a person without causing serious harm.

These features enable a collaborative work scenario, but having them doesn’t automatically transform a robot into some inherently benign creature. Even a robot marketed as collaborative can be dangerous if it wields a sharp tool or heavy part without proper safeguards. Conversely, a traditionally powerful robot might work collaboratively if the application around it (sensors, safety systems, and task design) ensures human safety. The new standards language recognizes this reality. By dropping the robot-centric term, the focus shifts to how the robot is used. “Human-robot collaboration relates to the application and not to the robot alone,” the standards writers emphasize. Hence, the pivot to “collaborative application” – a subtle but crucial change in phrasing.

It’s the Application That’s Collaborative (Not the Robot)

Under the updated definitions, collaboration is a property of the application – meaning the overall system in which humans and robots interact. This clarification is more than mere wordplay; it clarifies responsibility and scope. An application can be designed, validated, and confirmed as collaborative by meeting stringent safety requirements. A robot, by itself, cannot. From a standards perspective, you can only call a process collaborative after ensuring every aspect of the human-robot interaction is safe by design. This includes risk assessments, safety functions, and protective measures around the robot and its workpiece or tool.

Industrial robot safety experts have been nudging the industry toward this view for years. A 2021 analysis of collaborative robotics bluntly noted the widespread “misconception that collaborative robot = safe robot.” This misinterpretation, the authors argued, arose from the very term collaborative robot itself. As they explain, “robots alone are not collaborative, only applications can be rendered as collaborative when they also adopt sufficient safety features.” The phrase “collaborative robot” is “quite inaccurate” because it suggests a machine is safe in all situations, which is “usually not the case” without a proper risk analysis. The new ISO/ANSI terminology is essentially an official echo of this point: a robot isn’t magically safe or unsafe on its own; it all depends on the context and controls of its use.

Bill Edwards, vice-chair of the R15.06 committee, put it succinctly: “It is the entire collaborative application that must be safe.” This means engineers and integrators must consider the whole system – the robot, the task, the tools, the workspace, and the human operators – when implementing a collaborative scenario. By re-framing the terminology, the standards drive home that safety is a system-level attribute. The robot is just one component in a larger, collaborative puzzle.

Implications for the Robotics Industry

This shift in definitions may feel like a semantic nuance, but it has practical implications for how robots are marketed, selected, and deployed. For robot manufacturers, the message is clear: you can no longer simply sell a robot arm as a “collaborative robot” without context. Instead, you need to communicate what collaborative features the robot provides (e.g. force-limited joints, sensors), and guide customers on how to use them in a safe application. In fact, the updated standards have folded in guidance from the previously separate ISO/TS 15066 (the technical specification on collaborative robotics) directly into ISO 10218’s requirements for collaborative applications. This underscores that those “cobot” features are just part of a bigger safety strategy.

For system integrators and end-users, the new language serves as a reminder that compliance isn’t achieved by simply purchasing a specific type of robot. The entire workcell design must be assessed. The revised ISO 10218-2:2025 (for robot integration) even shifts terminology to emphasize robot applications rather than just robot systems. It expands the concept of a “safeguarded space” beyond fixed fences to include dynamic protective measures (scanners, sensors, etc.) suitable for both collaborative and conventional setups. And terms like “safety-rated monitored stop” – once associated mainly with cobot modes – have been renamed to “monitored standstill,” reflecting that such safety functions apply to more than just collaborative scenarios. All of these changes are designed to make standards more explicit and user-friendly, so that engineers don’t treat collaborative robotics as a mystical category requiring separate lore. It’s all just robotics, subject to the same rigor of risk assessment.

Perhaps most importantly, the industry needs to shed the myth that just because a robot has a friendly name (or a shiny force sensor) it can do no harm. Safety comes from careful design and testing of the application. The new ISO and ANSI standards are effectively telling everyone: Focus on what you’re doing with the robot, not what sticker is on it. This is important not just for accuracy, but for accident prevention. As one detailed study observed, a “false sense of safety” can arise if workers assume a so-called cobot will never hurt them. That complacency can be dangerous. By emphasizing collaborative applications, the standards encourage a more vigilant mindset: even when using the latest “safe” robot models, engineers must still verify the entire operation to ensure safety.

A Welcome Change (Even if Marketing Must Catch Up)

In an industry enamored with the term cobot, this change might ruffle some marketing departments. (After all, “collaborative application robot” doesn’t roll off the tongue quite the same way.) But among safety engineers and standards professionals, it’s a welcome clarification. It aligns the language with what they’ve been saying all along: true collaboration is achieved by design, not by decree. The updated ISO 10218 and ANSI/A3 R15.06 standards essentially formalize what should be common sense – a robot isn’t collaborative in isolation. Only when humans and robots successfully share tasks in a controlled environment do we have a collaborative robot application.

By dropping the misleading term, standards developers hope to improve communication and compliance. The robotics community is being encouraged to focus less on specialized “cobots” and more on safe collaborative solutions. This encourages broader thinking: any robot, big or small, can work with people if you apply the right safety measures, and conversely, even a lightweight robot can be hazardous without them. The emphasis on application also places responsibility squarely on integrators and users to do their due diligence. As the revised standard itself stresses, only a well-designed use of the robot – “the actual use of the robot” – can be deemed collaborative.

In sum, the industry stakeholder members at ISO and A3/ANSI have called time-out on the term “collaborative robot”. It served its purpose in raising awareness of human-friendly robots, but it also caused confusion by implying that safety was a built-in feature of the robot. The new standards make a pivotal correction: collaboration is not a robot product category – it’s an achievement of risk mitigation and smart integration. For the robotics industry, that’s an important distinction. It’s a sign of a maturing field that we’re moving past the hype and naming conventions, and focusing on what really matters: designing robotic applications where humans can work safely and productively with our mechanical colleagues.

---

Robot News Of The Week

Kodiak Robotics delivers its first factory-made autonomous truck

Kodiak Robotics has delivered its first autonomous truck, built directly off a factory production line, in partnership with Roush Industries. This partnership has established a dedicated line for Kodiak Driver-equipped trucks in Michigan. The truck was delivered to Atlas Energy Solutions, which ordered 100 vehicles in 2024 and has now received eight.

The partnership aims to scale production into the hundreds by 2026. Kodiak, which announced plans to go public via SPAC with Ares Acquisition Corp II, sees this milestone as proof the future of freight is arriving at commercial scale.

Zoox bets big, launches robotaxi service on Vegas Strip

Amazon-owned Zoox has debuted its custom-built, fully driverless robotaxis on and around the Las Vegas Strip. Riders can book free trips via the Zoox app to destinations like Resorts World, AREA15, and Topgolf, with dedicated pickup zones and concierges at select spots.

Unlike retrofitted cars, Zoox’s vehicles are purpose-built—no steering wheel or pedals, with face-to-face seating for a shared ride experience. The company says it can produce up to 10,000 units annually and plans to expand to Austin and Miami after testing in San Francisco and Foster City.

Zoox enters a growing U.S. robotaxi field that includes Waymo (over 100M autonomous miles), Nuro, and Tesla’s pilot in Austin.

Universal Robots unveils UR8 Long cobot for space-constrained automation tasks

Universal Robots introduced the UR8 Long, a new cobot designed for demanding automation tasks, now available for order with shipping in October.

With a 1,750 mm reach (same as the UR20) but a slimmer, lighter build (30% less mass), the UR8 Long supports an 8 kg payload, making it ideal for space-constrained setups, welding, bin picking, and precision inspections.

The cobot runs on PolyScope 5 & X and can be extended with MotionPlus for smoother, more accurate trajectories. Its compact wrist and intuitive lead-to-teach programming allow quick setup, while advanced motion control and upgraded freedrive boost flexibility.

UR says the UR8 Long delivers faster cycles (up to 30% improvement), higher weld quality, and expanded bin-picking reach—helping manufacturers save time, cut rework, and attract new talent.

---

Robot Research In The News

Artificial intelligence enables exoskeletons to assist users more efficiently

Researchers from Japan’s RIKEN Guardian Robot Project have developed an AI-driven exoskeleton that adapts to both the user’s condition and environment, overcoming limitations of traditional systems that rely on pre-programmed motions or muscle sensors (EMG).

Using a transformer model with inputs from a camera near the user’s eyes and body-mounted sensors, the system successfully supported tasks like picking up objects and climbing steps—reducing muscle strain and improving mobility.

Crucially, the assistive strategy generalized across different users without retraining, a major advance for exoskeleton tech. The team sees applications in healthcare, rehab, and eldercare, with the potential for personalized, adaptive support that boosts independence and quality of life.

---

Robot Workforce Story Of The Week

Indiana University Launches New Robotics Degree to Meet Workforce Demand

Indiana University Bloomington has approved a Bachelor of Science in Robotics, to be offered through the Luddy School of Informatics, Computing and Engineering. Backed by IU’s $75 million investment in human-centered engineering, the program will train students in AI, programming, design, and control systems, while integrating industry experience and entrepreneurship.

The degree is designed to meet growing workforce needs in robotics, automation, and advanced manufacturing, preparing graduates for immediate career placement and supporting innovation, technological advancement, and economic growth in Indiana and beyond.

---

Robot Video Of The Week

Dusty Robotics is transforming construction layout with its robotic field printer, which transfers digital floor plans directly onto building sites. Instead of surveyors using tape measures and chalk lines, the robot autonomously navigates the floor and prints walls, doors, outlets, and plumbing lines with millimeter accuracy. This speeds up layout from days to hours, eliminates costly errors, and ensures every trade works from the same blueprint. By linking directly to BIM files, Dusty’s system bridges design and construction, reducing rework and delays. It’s a clear example of how robotics can address labor shortages and boost productivity on-site.

---

Season two of The Following Seas podcast begins with guest host Aaron Prather talking with Barbara Bell, Ed.D., CAPT US Navy (ret), about lessons learned both in the cockpit and classroom.

In this latest episode, Captain Bell shares her remarkable journey — from entering the U.S. Naval Academy to serving as a naval aviator and flight test officer — and learn how she is shaping the next generation as a professor of leadership and ethics at the United States Naval Academy.

Together, Prather and Bell discuss how navigating turbulence — both in flight and in life — can inspire stronger, more resilient leaders.

Listen wherever you get your podcasts, or visit https://lnkd.in/eegHK7kJ

---

Upcoming Robot Events

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)

Oct. 29-31 Intl. Symposium on Safety, Security, and Rescue Robotics (Galway, Ireland)

Nov. 3-5 Intl. Robot Safety Conference (Houston, TX)

Dec. 1-4 Intl. Conference on Space Robotics (Sendai, Japan)

Dec. 11-12 Humanoid Summit (Silicon Valley TBA)

Mar. 16-19 Intl. Conference on Human-Robot Interaction (Edinburgh, Scotland)

Mar. 29-Apr. 1 IEEE Haptics Symposium (Reno, NV)

---