That Kraft employee I blogged about who was killed by an AGV a few weeks ago is still on my mind. I wasn't kidding when I said the accident surprised me, considering how fail-safe I believed this technology to be. My being surprised about this is one thing, but when a guy like Roger Bostelman is surprised by anything involving industrial vehicle technology, you know this was an unusual event.
Roger is an engineering project manager for the National Institute of Standards and Technology's (NIST) Intelligent Systems Division, and he and a team of colleagues have been researching ways to improve AGV safety standards. He sent me an e-mail after reading my blog and expressed his shock at what happened. “Shock” was his word, not mine.
“Looks like we'll keep rolling to make AGVs and forklifts as close to 100% safe as best we can,” he wrote.
Turns out, that's his current focus in life. He cited one section of my blog where I excerpted a paragraph from the existing ANSI/ITSDF B56.5-2012 safety standard for AGVs that I suggested be printed in bold type:
“Although the vehicle braking system may be performing correctly and as designed, it cannot be expected to function as designed and specified should an object suddenly appear in the path of the vehicle and within the designed safe stopping distance. Examples include, but are not limited to, an object falling from overhead or a pedestrian stepping into the path of a vehicle at the last instant.”
Bostelman's research at NIST is focused on Safe Control of Mobile Autonomous Vehicles used for Manufacturing. His team is developing standard test methods to measure reduced vehicle energy when in the “exception,” or “stop” zone.
“Although there should be AGV user/supplier agreements in place for this, we feel there could still be an issue,” Bostelman wrote. “We're also studying prediction of when people are headed towards the AGV path unseen by the optic safety sensor towards reducing vehicle energy there. This too, could occur within the exception zone.”
They're basically looking at several advanced methods to measure operator visibility that aren't included in the ANSI standard. One of the technologies AGV manufacturers should pay attention to, particularly in light of this fatality, is 3D imaging sensors. These can detect static and dynamic objects, on the ground or overhanging, that are in the path of a vehicle and then enable the vehicle to stop without hitting the object.
These sensors are not safety rated, but Bostelman's group is recommending that the ANSI/ITSDF B56.5 standard committee develop test methods that can be used by all AGV manufacturers and users in anticipation of when these sensors are safety rated.
These are some of the modifications NIST is recommending to the ANSI/ITSDF B56.5 standard committee if they include 3D sensors in the revised document:
• Obstacles that enter the exception (stop) zone of the AGV must be detected and it must be shown that the vehicle energy is reduced to a safe level as determined by the committee;
• Although 3D sensors are not yet safety rated, AGV non-contact safety sensors may be augmented by 3D sensors to provide improved obstacle detection;
I welcome input from any AGV vendor or user on this so I can share it with Bostelman—whether research or experience related. Even though that “Death by AGV” story was as unusual as a “Man Bites Dog” headline, there are probably many more incidents where a fatality was narrowly averted, either by virtue of technology or of sheer providence. These are rarely reported. It's time they are. We may be able to make a huge contribution to occupational safety—even if it saves one life.