Robotics and automation are often mentioned together, notably in the name of IEEE RAS, the IEEE’s robotics and automation society. Following a meeting of its communications committee, IEEE experts informally discussed what these two words cover, and the importance of the distinction for the community. Their reflections do not settle the question, but they highlight the stakes.
- Automation makes processes repeatable, efficient, and predictable, often without being noticed.
- Robotics requires a physical body and is also studied for its own sake.
- The boundary between robotics and automation stays blurry, as physical AI raises the same question.
- Inconsistent definitions across countries distort statistics and national comparisons of robot numbers.
For Mariateresa Pedone, automation is almost invisible in daily life: it mainly serves to make processes repeatable, efficient, scalable, and predictable. Robotics becomes relevant when automation requires a physical presence. David Garzon Ramos expands on this view by noting that automation has primarily served to support other activities, while robotics has also developed for its own sake, to understand the principles of intelligence, control, and embodiment (embodiment, meaning acting through a physical body).
Blurring Boundaries
The question of whether robotics is a subcategory of automation divides less than it prompts reflection. David Garzon Ramos believes no one necessarily wants to say one field is broader than the other, and he notes the same question arises today with physical AI (physical AI). Claudio Pacchierotti points out that not all robots are automated: a robot can be fully teleoperated, meaning remotely controlled by a human. Conversely, automation does not necessarily require a robot and can be entirely software-based.
Everyday examples further complicate the picture. Ellen Rumley mentions a large factory with a computer-controlled conveyor system that sorts packages: should the whole system be considered a robot, even though it is not a unit that moves or seems even slightly alive? Claudio Pacchierotti himself uses a simple question when teaching robotics to elementary school children: ‘Is it a robot?’ A washing machine physically transforms laundry and is highly automated, but most people do not see it as a robot. Amy Eguchi, whose background is in education, observes in Japan increasingly automated household appliances and products marketed as ‘AI microwaves.’ She wonders if it’s a robot or just automation.
Concrete Consequences
These definitions have effects that go beyond linguistics. Ellen Rumley, who compares industrial robotics and service robotics between countries, finds that categories are not always consistent: robot vacuum cleaners, for example, can be counted differently depending on the country. She explains that China may appear to produce millions more service robots, partly due to what is included in the category. Sabine Hauert concludes that the issue also touches on statistics, national comparisons, and the perception of a country’s technological strength.
Language also evolves with research trends. David Garzon Ramos says he now uses terms like ‘physical AI’ and ’embodiment’ in his projects, while still introducing himself as a roboticist to his peers. For Sabine Hauert, this complicates communication, as every new project seems to need to mention robotics, automation, and sometimes physical AI. David Garzon Ramos believes vocabulary must be adapted to the audience: with children, ‘robot’ immediately sparks interest, whereas ‘automation process’ does not. Sabine Hauert finally notes that the names of conferences and professional communities preserve a trace of the history of these fields, ‘closely related fields.’
