Where does the 41 robot deaths figure come from?

A research paper by a NIOSH scientist, published in a peer-reviewed journal that sits behind a paywall, drawing on a restricted federal fatality research file. That is why it is so hard to verify or reconcile with other counts.

Does OSHA publish a robot death count?

No. OSHA publishes robot-safety guidance and individual case narratives, not an aggregate fatality figure. Its Technical Manual gives checkable safety facts like the 250 mm/second teaching speed, but not a death toll.

Robot safety

Robot workplace deaths: what the public record does and doesn't show

It sounds like a simple question with a simple answer: how many people have been killed by industrial robots? The figure you will most often see is about 41 deaths in the United States between 1992 and 2017. Try to trace it to a source you can actually open, and it falls apart.

The paper behind the number sits in a paywalled journal. The government agency that summarized it blocks automated readers, and that same agency has published a different total in the past. The underlying federal injury data is restricted. What is left, that anyone can fetch and check, is guidance, not a body count.

So this page does the honest version. It traces what is documentable about robot workplace safety to a source you can open, and it is plain about where the widely-quoted death toll cannot be verified.

Data covers robot-related fatality claims, 1992 to 2017, against OSHA robot-safety guidance. Last reviewed by a human editor before publication.

The figures and where they come from

Each figure is rated for how safely you can cite it today. Ratings judge current usability, not whether a number was ever correct.

Figure	What it is	Source	Citation Confidence	Notes
2 million+ robots	Robots in workplaces worldwide (2018)	[A]	High	OSHA, citing the IFR, on the installed base as of 2018. Context for the exposure, and a figure you can open and check.
40,000+ installed	US robot installations (2018)	[A]	High	Robots installed in the United States in 2018, per OSHA citing the IFR.
250 mm/second	Max robot speed during teaching	[A]	High	The reduced speed OSHA specifies for teaching, 10 inches per second or 250 mm/second or less. A concrete, citable safety figure.

Why the numbers disagree

There is no single agreed robot death toll because no open, primary source publishes one. The most-cited figure, about 41 deaths from 1992 to 2017, comes from a research paper by a NIOSH scientist published in a peer-reviewed journal. That journal is paywalled, the government summary of it returns an access error to automated readers, and the underlying federal fatality data is a restricted research file, not a public table.

The numbers are not even consistent. Earlier NIOSH statements referenced a higher count for an overlapping period, so the same agency has effectively published two different totals depending on when and where you look. Without an openable primary, there is no way to reconcile them.

What OSHA does publish is guidance, not a tally. Its Technical Manual describes robot hazards, specifies safe teaching speeds, and notes that most robot accidents happen during setup and maintenance rather than normal operation. Those are checkable facts. The death count is not one of them.

How to cite these figures

If you need to say something concrete about robot safety, cite the OSHA Technical Manual: the reduced teaching speed of 250 mm/second, the note that even 3-kilogram robots can be dangerous, and the finding that accidents cluster during assembly, installation, and testing. Those are open and checkable.

If you must reference the death toll, say plainly that the figure of about 41 traces to a paywalled paper and cannot be verified through a free primary source, and that agency counts have not been consistent. That caveat is more accurate than the number alone.

Do not present any single robot-death figure as authoritative. The honest statement is that the United States does not publish an open, current, verifiable count of robot-related workplace deaths.

Where people go wrong

Quoting about 41 robot deaths as a settled government statistic. It comes from one paywalled paper, and the agency behind it has referenced different totals, so it is neither open nor consistent.

Attributing the figure vaguely to OSHA or the BLS. OSHA publishes robot-safety guidance and individual case narratives, not a fatality count, and the underlying BLS fatality research file is restricted.

Treating the absence of a clean number as proof that robots are safe, or as proof they are dangerous. The absence is a data-availability problem, not evidence either way.

How we checked

The checkable figures on this page come from the OSHA Technical Manual, Section IV, Chapter 4, which we fetched and confirmed contains each figure: the 2-million installed base, the 40,000 US installations, the 250 mm/second teaching speed, and the 3-kilogram note.

For the death toll itself, we looked and could not find a free, primary, machine-verifiable source. The peer-reviewed paper is paywalled. The government pages that summarize it return an access error to an automated fetch. The underlying federal fatality data is a restricted research file. We are reporting that absence rather than laundering the number through a secondary blog.

This is deliberately a page about what is documentable. Where a figure could be opened and checked, we cite it. Where the most-repeated figure could not be, we say so, because on a claim-tracing page the sentence that matters most is the one that admits the number does not trace.

Full source list

Primary sources, with live links. Every figure above traces to one of these.

[A]U.S. Occupational Safety and Health AdministrationAccessed July 14, 2026
OSHA Technical Manual (TED 01-00-015), Section IV, Chapter 4, "Industrial Robot Systems and Industrial Robot System Safety"
https://www.osha.gov/otm/section-4-safety-hazards/chapter-4

Common questions

How many people have been killed by industrial robots?: There is no open, verifiable count. The most-cited US figure is about 41 deaths from 1992 to 2017, but it comes from a paywalled paper, the government summary blocks automated access, and the same agency has referenced different totals, so it cannot be confirmed through a free primary source.
Where does the 41 robot deaths figure come from?: A research paper by a NIOSH scientist, published in a peer-reviewed journal that sits behind a paywall, drawing on a restricted federal fatality research file. That is why it is so hard to verify or reconcile with other counts.
Does OSHA publish a robot death count?: No. OSHA publishes robot-safety guidance and individual case narratives, not an aggregate fatality figure. Its Technical Manual gives checkable safety facts like the 250 mm/second teaching speed, but not a death toll.
What can you actually cite about robot safety?: The OSHA Technical Manual: the reduced teaching speed of 250 mm/second, the note that even 3-kilogram robots can be dangerous, and the finding that most robot accidents happen during setup and maintenance, not normal operation.

More data, traced to source

Robot reliability numbers: the vendor claims and the one independent study
Manufacturers advertise robot uptime in the high nineties and mean time between failures in the tens of thousands of hours. The one independent study of more than 400 factories found a robot cell is reliable 88 percent of the time, with 87 minutes between failures.
Robot density statistics and where they come from
The most-cited robot density figures are stale. Here are the current numbers, where each one comes from, and why the global average fell while automation kept rising.