A group of Silicon Valley roboticists who developed Shakey, a pioneer mobile robot project, gathered last night at the Computer History Museum in Mountain View, Calif., to dedicate the tall, wheeled machine as an IEEE Milestone. Joining the group were other robotics visionaries, IEEE officers and local IEEE section members, and fans of computing history. Shakey, developed at SRI International between 1966 and 1972, was honored as the world’s first mobile, intelligent robot.
The official plaque, marking Shakey as an IEEE Milestone, reads:
“Stanford Research Institute’s Artificial Intelligence Center developed the world’s first mobile, intelligent robot, SHAKEY. It could perceive its surroundings, infer implicit facts from explicit ones, create plans, recover from errors in plan execution, and communicate using ordinary English. SHAKEY’s software architecture, computer vision, and methods for navigation and planning proved seminal in robotics and in the design of web servers, automobiles, factories, video games, and Mars rovers.”
Although Shakey itself didn’t take the podium—the historic robot now resides in a glass case on the exhibit floor of the Computer History Museum—the event closely mirrored an award ceremony honoring a human’s illustrious career. Dignitaries read commendations; former colleagues reminisced, telling a few tales that might not previously have been presented for public consumption; and a “grandchild” made a guest appearance.
Computer History Museum chairman Len Shustek kicked off the evening, recalling his days at Stanford when “we all knew about Shakey.” He added, “He was a legend to us.” Then Savioke’s Relay delivery robot (Shakey’s “grandchild” robot, whose navigation system relies on the A* algorithm used to direct Shakey) rolled into the room and presented the evening’s agenda to IEEE Milestone coordinator and event organizer Brian Berg, and the reminiscences (coordinated by New York Times writer turned Computer History Museum historian John Markoff) began. Here are some highlights:
On the goals of the Shakey project:
“It was supposed to be an ‘automaton for reconnaissance.’ We thought it would have something to do to the military,” said Peter Hart, one of the Shakey project members and most recently senior vice president of Ricoh.
“Someone had in mind dropping the robot on a road in China and have it count tanks,” said Nils Nilsson, another Shakey project alum and Stanford computer science professor emeritus.
On why Shakey had no arms:
“The ground rule was to keep it as mechanically simple as possible; that’s why we don’t have arms [and why we] didn’t miniaturize components. It looks like a washing machine on wheels,” said Hart.
“I thought we could do most of the planning and manipulate objects by pushing them without an arm,” said Nilsson.
On getting—and keeping—the project funded:
“The first contract—about 18 months—was about $300,000. We did get renewed through ’71, then DARPA said ‘no more robots,’ ” Nilsson explained.
“That was because of the ‘Mansfield amendment,’ ” interjected Hart, adding, “It said DoD research had to be directly militarily related.”
“After DARPA quit funding robotics,” Nilsson continued, “we still wanted to do the software. So [we decided to] have a human take the place of a robot, [and we’d] give instructions to the human and the human could follow them and do things. That kept us going for a couple of years, then DARPA saw through that.”
During that era of government contracts, recalled Hart, “a U.S. auditor stopped by my office and started quizzing me. First thing he asked: ‘Had I taken delivery of 8 billion packets of bits?’ He read out an exact 10-digit number. So I’m polite, I say that sounds about right, and he checks it off. Then he asked if I had set up procedures to inspect the condition of incoming packets. I thought, ‘There must be an error-detecting code,’ so I said yes to that, too. And he checked it off. Then he wanted to know if the packets had arrived in good condition. ‘Dr. Hart,’ he asked, ‘was there any tarnish or corrosion on the packets?’ This I could reassure him, no corrosion. Then he asked if I had adequate warehousing facilities. I was tempted to say, ‘Didn’t you see all those disk drives? The packets are in there, across the hall.’ But he didn’t have much of a sense of humor, so I just said yes, and he seemed satisfied.”
On the prescience of union representatives:
“Union people came in to visit us,” said Hart. “They were wondering if robotics and automation would disrupt employment. And in the long run they were the most accurate.”
On the development of the seminal pathfinding algorithm A*:
“A* is the way of computing the shortest route between A and B,” Hart said. “It had two nice properties: We could prove it always works, and also prove that it does the minimum amount of computation. Bert [Raphael] and Nils [Nilsson] cooked this thing up between them, and I came in and thought we could prove it. But then we couldn’t get it accepted by any reputable journal of the day. Bert was so irritated by that, he went off and founded another AI journal [the Journal of Artificial Intelligence] just so he could publish it. Eventually it got published in an IEEE Transactions.”
On strange behavior:
“Shakey would sometimes stop what it was doing and start pivoting around 360 degrees around and around. Finally we dug down in the code and found that there was a routine in there intended to unwind cable,” said Hart. It had taken a while to get FCC approval for a radio antenna to make the robot wireless, so, initially “we had a giant cable to the ceiling.”
On Shakey’s legacy:
“There were individual algorithms and architectures,” said Hart, “but in my opinion the legacy was the totality of Shakey. It was the first integrated system that could perceive, reason, and act. It gave confidence and inspiration to people that they could do this, and they could do a lot better.”
On robotics today:
“I’d like to see more integration” of AI and robotics, said Nilsson. “Today’s robots can wander around well—the two-legged ones can stand up, they have good control systems—but they don’t know what they’re doing.”
On the robotic exploration of space:
“When I think about the astronauts that went to the moon, they took a rock hammer and broke open rocks,” said JPL’s Matt Heverly, who served as lead driver for the Mars Curiosity rover. “The hammer was a tool. Now, the robots aren’t doing the exploration; humans are. The robot is just a fancy rock hammer.”
“We worked on remote presence robots,” said Steve Cousins, founder of robot maker Savioke and former CEO of Willow Garage, the legendary Silicon Valley robotics lab. Remote presence “is great,” he says. “I just did one in Japan, but I wasn’t there, I didn’t have any sushi. It really is different to experience it. Unless we send a human to Mars, all we’ll get back is video.”
“If you want to do great science remotely, send a robot,” said William Mark, head of SRI’s information and computing sciences division. “But there is the inspiration, the thrill, the vicarious experience of astronauts that is going to be hard to replicate in a robot.”
On Shakey’s grandchild Relay:
“We started by asking, ‘Can you put a robot around people?’ ” said Cousins. “We have flat terrain, not a Mars-like terrain problem. But we have another challenge; we put robots in hotels, where you have drunk people. That is another obstacle you have to deal with. We made the robot cute, so even drunk guys, when they knock it over, usually pick it up.”
Relay’s designers, Cousins said, intentionally chose not to make it talk. “We just have beeps and whistles like R2-D2—that helps set expectations. If we make it talk, people will say, ‘Bring me a gin and tonic.’ It can’t do that.”
“We made this amazing robot at Willow Garage,” he told the audience, referring to the PR2, a mobile manipulation robot. “It had arms, it could do anything. People would ask ‘What would it do?’ and we said, ‘Well, anything.’ That wasn’t a good answer. And then they would ask, ‘How much does it cost?’ and we would say, ‘$400,000.’ ”
“My goal was to build something that did something that is worth more than it costs,” Cousins said of Relay.
The IEEE Milestone program, launched in 1983, honors significant inventions, locations, or events related to electrical engineering and computing that have benefited humanity, and which are at least 25 years old. IEEE President-Elect Jim Jefferies, speaking at the event, pointed out that this dedication marks the 174th such piece of engineering history honored, and is the 12th for the IEEE Santa Clara Valley Section.