I’m going to switch gears today and talk about industrial automation.
Last December, Rockwell Automation – one of the leading companies in industrial automation – released a white paper entitled “International Safety Standards Keep Pace with Advances in Robotic Technology and Applications”. This paper talks about the current robotics safety standard (ISO 10218-1 and ANSI/RIA R15.06), and the new standard (part 2 of ISO 10218, or ISO 10218-2) being drafted for release in 2011.
With a brief overview of history of robots, the paper talks about how ANSI/RIA R.15.06 adopted in 1999 to address robot safety issues does not adequately cover newest innovations in the field. The new standard will cover standards for cableless teach pendants, human-robot collaboration, robot-to-robot synchronization and vision-based safeguarding systems.
From my limited experience and exposure in the automation industry, I was taught two golden rules to building a safe automation system: implementing Poka-Yoke system, or put barriers between the person and the machine/robot (e.g., work cells). As the paper points out, such old way of ‘safely’ interacting with a robot in a human-robot collaborative task requires safety control system, which limits robot motion and decreases productivity. New software-based system with allows people to share the same workspace with lower risk.
While discussing the topic of safety, he talks about how the current standard for industrial robots instructs us to always physically separate humans and robots. He advocates that increasing the transparency of a robot’s intention to act by means of gestures, voice and context can be a practical step towards considering ethics in robot design.
It is unfortunate that due to the nature of the robotics field, a decade old standard is not adequate, and the upgrade to this old standard (to be released next year) still doesn’t sound too promising in covering all of the newest innovation in the field. If technology is developing at an exponential rate, does that mean we need to have exponentially faster standard updating system as well?
But one thing seems obvious. Before the robots become our nannies, husbands/wives, and even children, robot safety issues surrounding currently existing robots need to be addressed – they’re more urgent in a way because these robots exist in multitudes already.