After decades of promises of autonomous machines, robotics is finally poised to transform life and business as we know it, from manufacturing to elder care.

That’s the consensus of scientific researchers, entrepreneurs and corporate executives who gathered this week at the TechCrunch Robotics conference in Cambridge, Massachusetts. If there was an overriding message from the daylong dialogue at the Massachusetts Institute of Technology this week, it was that new technologies such as machine learning coupled with innovative applications means the robot revolution is nearly here.

“We’re now in the golden age of robotics,” declared Sami Atiya, president of robotics and motion for ABB Ltd., a global supplier of industrial automation technology. “If you think this is going fast, then fasten your seat belts, because it’s going to go faster.”

MIT showcases robotics research

MIT’s Computer Science and Artificial Intelligence Laboratory does not often let the public see what is transpiring behind the research curtain, so the school’s active participation in the TechCrunch agenda ended up revealing some intriguing possibilities. Daniela Rus (pictured below), director of CSAIL, described how major advances in 3D printing technology were allowing researchers to now print “origami” robots, designed using a print-and-fold fabrication technique. Attach a motor and small electronics and off they go.

“We hope that in the near future, we’ll have robots that literally walk themselves out of the printer,” said Rus.

This technique now offers the opportunity to create robots in miniature, small enough to be deployed inside the human body and eliminate the need for surgery. Rus showed video footage of a test demonstration inside artificial organs, where a tiny robot was ingested in an ice cube. After the ice melted, the robot was released and it successfully found and removed a button battery in the stomach.

MIT Professor Sangbae Kim showcased the Cheetah, an animal-like robot that can run at 13 miles per hour, navigate balanced turning, stretch and even shake hands on command using an Amazon Alexa natural-language programming interface. Through Kim’s experimentation with structural and motor technology in MIT’s Biomimetics Robotics Lab, the Cheetah can jump and land safely over multiple hurdles, a significant advancement by itself.

MIT’s Daniela Rus (Photo: Paul Marotta/Getty Images)

MIT researchers were joined at the conference by a number of corporate representatives as well, including Amazon.com Inc. The company has built its reputation on cloud computing and global e-commerce, but Amazon has also made a sizable investment in autonomous technology, using 80,000 robots in its fulfillment centers.

Amazon and the picking challenge

At the heart of Amazon’s robotics work is picking, the ability for a robot to correctly find, grasp and deliver a warehouse item for shipping. This is an important, evolving field for industry, one that combines the mechanics of autonomous manipulation, or grasping, with the ability for a robot to properly recognize what it is supposed to pick up.

“We have seen some success in the picking challenge,” said Tye Brady, chief technologist at Amazon Robotics, but declined to elaborate further on what the company is developing. Amazon is holding a “Robotics Challenge” later in July that is specifically geared towards advancements in the picking task.

Another company with ambitious plans for robotics is Toyota, and it’s not just centered on the self-driving car. The Japanese carmaker recently invested $14 million in the startup Intuition Robotics Ltd., which is developing a robot called ElliQ to deliver elder care services.

“Robots can make a tremendous difference,” said Gill Pratt, chief executive of the Toyota Research Institute. “People have an ability to empathize with machines.” Pratt also noted that he didn’t think it would be that difficult “to build machines that can track how we feel too.”

Pratt was more cautious about committing to a definitive time frame when autonomous cars would become ubiquitous, citing the need for more training data. “It may be that writing software is not the key technology,” said the Toyota executive. “The answer may be in the volume of the data that’s necessary to collect to make these cars better.”

Robots are not merely confined to the ground. Intel Corp. has been showcasing what they call the “drone light show,” an aerial fleet of 300 synchronized, lighted drones making shapes and figures in the night sky. The company, which is making its own commercial drones dubbed Falcon 8+, gave over 100,000 concertgoers at the Coachella Valley Music Festival in California a sneak peek at the “drone fleet” concept earlier this year.

Intel’s Falcon 8+ drone (Photo: Intel)

Jan Stumpf, head of drone architecture for Intel, told the gathering that he expected “to see a lot more innovation” in the handling and use of data for drone technology. This could potentially include using coordinated drone fleets, like the one at Coachella, for search and rescue operations in forests. “That could be very powerful,” he said.

Roboticists are clearly focusing much of their attention these days on artificial intelligence and how to make robots smarter. The field is running on two parallel development tracks, one focused on the mechanical and the other on cognitive. “In the near future, cognitive tasks and physical tasks will be equally transformed by AI and robotics,” said Rus.

Robots learn from other robots

How might AI help evolve robots? Claudia Perez-D’Arpino, a Ph.D. candidate at MIT, presented a new coding system called C-LEARN which is designed to improve robotic learning with less programming. The idea is to give robots data and then show them what’s supposed to happen.

This knowledge base is designed to speed up the learning process, but it also comes with a significant bonus. Robots can pass this learning onto other robots, a transfer that Perez-D’Arpino said she has successfully accomplished in the MIT lab.

Advances in machine learning like this are making some people nervous, even alarmed. One of them is Tesla Inc. founder and Chief Executive Elon Musk, who recently called for government regulation in the field to avoid a robot takeover of the human species.

Several of the conference participants commented on Musk’s warning, which he repeated this week, and none took his side. “I’m not a fan of regulation,” said Amazon’s Brady. Toyota’s Pratt also said that he was “skeptical” about Musk’s concerns over the social disruption of AI.

Rodney Brooks, who co-founded iRobot and is currently the chairman and chief technology officer of Rethink Robotics Inc. (Sawyer and Baxter models pictured, above), also disagreed with Musk, and pointed out that people who believe that AI needs regulation mistakenly assume that just because a technology is successful in one field, it can translate that success into others. “By the way, let’s talk about regulation on these self-driving systems for Tesla,” said Brooks.

Musk’s recent comments and the reaction that followed reinforce the growing belief that issues surrounding AI are getting more attention because the speed of robotic innovation is accelerating. It has happened with other technologies such as smartphones, where the concept bubbles along quietly and then suddenly everyone can’t live without it.

“My great aspiration is to have a world where anyone can use a robot,” Rus said. More than ever, it looks like she may not be that far away from realizing her goal.

Photo: Rethink Robotics