Late in the last decade, when I was a senior editor at Control Engineering magazine, one of the biggest worries in the robotics world was safety. Robots were viewed like the proverbial bull in a china shop.
Whether fixed or mobile, the things were considered big, clumsy and dumb as a box of rocks. They had no sense of what was going on around them, so any time a human blundered into their work area, the human was risking serious injury. The robot would literally step on them without a second - or even first - thought.Another problem was that programming the things was as expensive - maybe even more expensive - than buying them in the first place. Just figuring out how to tell the slug-brained idiots what to do was a major project for highly trained and skilled programmers.
Seeing this state of affairs motivated me to fantasize about what working with an ideal robot would be like. In my Red McKenna novels, I described such a technology, which I called “verbal programming.” It upended the way people worked with robots. More importantly, it upended the requirements for who could be trusted to operate them.
In “Down the Rabbit Hole,” maverick millionaire technical genius Doc Manchek, as a personal favor for his ex-girlfriend, recruits a stripper from a local nightclub to work for him as a robot programmer. It works in the story because of the new technical paradigm for the robots he was using.
When challenged about it, he said: “I didn’t hire a stripper. I hired a young woman who has good communication skills, above average intelligence, listens well, obeys instructions, asks questions when she doesn’t understand and shows motivation to join the team. Those are the main criteria for this particular position…”
That’s what we need it to be like in the future.
It was only possible in the story because the fictional robot was imbued with situational awareness and good communication skills. It was able to hold a two-way conversation with a human handler in spoken English. Some other capabilities it had, which would be necessary for any robot to work with an untrained human, included exquisite machine vision capable of building a mental model of the robot’s surroundings; compliant joints capable of being maneuvered by a handler attempting to teach the robot a new movement; and touch sensitivity so the robot could sense and respond to surprises.
I just ran across an industrial robot from a company called Rethink Robotics founded by the guy irresponsible for the Roomba autonomous vacuum cleaner. Named Baxter, the thing has most of the capabilities that made my fictional robot work well with an untrained human.
The only thing my fictional robot had that Baxter seems to lack was the ability to communicate via spoken English. Other systems have proven such capability is technically feasible, so we’re just waiting for someone to put it all together.
C.G. Masi has been blogging about technology and society since 2006. In a career spanning more than a quarter century, he has written more than 400 articles for scholarly and technical journals, and six novels dealing with automation’s place in technically advanced society. For more information, visit www.cgmasi.com.