Mind Matters Natural and Artificial Intelligence News and Analysis
Mensch und Roboter in der Medizin der Zukunft, Beruf Arzt am Arbeitsplatz

Can Robots That Work With People Ever Be Safe?

Robot IQ offers five reasons why not

Cobots are robots designed to be friendly to people. But some doubt that friendship will work:

It can be tempting to think of risk as an either/or situation — either your application is safe or it isn’t. In reality, risk is a sliding scale and you can never get rid of all risks completely. You can only know the true risk of a particular task by performing an adequate risk assessment. You need to do this whether the robot is collaborative or not.

Truth: cobot safety can be changed to suit task performance

The reality is that cobots have always been high-performance robots suitable for a range of industrial applications. Instead of being lesser robots, as some people mistakenly believe, cobots are often fully-fledged industrial robots with added functionality to make them safer.

These added safety limits include reduced speed mode, reduced force mode, added safety sensors (e.g. tactile sleeves around the robot to detect interactions), etc.

It is possible to deactivate these safety limits. This improves the performance of the robot, allowing it to move faster and apply more force to the task. Some experts believe that cobot safety is too high in the first place.

Alex Owen-Hill, “5 Reasons Cobots Will Never Be Completely Safe” at RobotIQ

This has the makings of a great sci-fi film.

What if they form a union?

Anyway, while we are here: The three laws of robotics have failed the robots

Mind Matters News

Breaking and noteworthy news from the exciting world of natural and artificial intelligence at MindMatters.ai.

Can Robots That Work With People Ever Be Safe?