The Future of Robotics and Artificial Intelligence Is Open

Openness is disruptive, profitable, and will allow participation in robotics and AI to take off

4 min read

This is a guest post by author William Hertling. The views expressed here are his own and do not reflect those of his employer, the IEEE, or IEEE Spectrum.

At South by Southwest Interactive last month, I debated the future of artificial intelligence with my co-panelists.

The roboticist on the panel argued that AI is an intellectually challenging field where the problems are difficult, and therefore can be solved only by highly intelligent people working on obscure mathematics and algorithms. The future, he argued, will look much like the past: a series of incremental, hard-won improvements in very narrow fields.

I disagree.

I'm sure, a decade ago, the folks at Encyclopedia Britannica would have believed the encyclopedia business was going to remain just the same. But the advent of Wikipedia proved that crowd-sourced knowledge curation provided broader, deeper, and more accurate results than the traditional small pool of experts.

Wired editor Chris Anderson, in his book The Long Tail, called this effect democratizing the tools of production.

Linux, developed by a community of volunteers, became competitive to established commercial alternatives in less than 10 years. Today, Linux is the de facto standard for servers and smartphones, where it is the foundation for Android.

open robotics Darwin-OP schematicRobotsource

In 2006, Netflix announced the Netflix Prize, a competition to develop better collaborative filtering algorithms. Within six days of the start, a team beat Netflix's own in-house algorithm. More than 2,000 teams ultimately participated, with hundreds of them besting the original algorithm.

Today, the Apache Project Mahout, created as a result of the Netflix Prize, is an open source collection of machine learning libraries. What once required specialized knowledge is now available in a library anyone can reuse.

By solving basic problems and making the solutions available for reuse as a platform, we enable newcomers to stand on the shoulders of giants and innovate at a higher scale.

Open platforms are disruptive.

The business of Encyclopedia Britannica suffered as a result of Wikipedia, but the entire world has benefited from open, free access to an even greater wealth of knowledge. Linux put a cramp in Microsoft's server operating system business, but the entire web economy, from Google to Amazon to Facebook, is powered by Linux.

The reason for these open platform successes is not because they were free (although free does increase adoption), but because the open platform produced a superior product.

Using closed-source as a form of economic protectionism for individual companies is not the solution. It just ensures competitors will advance faster.

The reason we haven't seen even greater amateur participation in robotics and AI, up until this point, has been because of the cost: whether it's the $400,000 to buy a PR2, or $3 million dollars to replicate IBM's Watson. This too is about to change.

IBM's Deep Blue beat Garry Kasparov in 1997 using 510 processors. But by 2006, the focus in chess playing algorithms had switched to software, with the best programs running on a dual processor computer.

IBM's Watson used 2,880 processing cores to win Jeopardy in 2011. Due to exponential growth in computational power (about 1.5x per year), by 2025 hobbyists will be able to run Watson-equivalents on a home PC.

One current example of this trend towards accessibility is ArduPilot, an autopilot created by the DIY Drones community for flying unmanned aerial vehicles.

I emailed Chris Anderson, founder of DIY Drones and author of the upcoming Makers: The New Industrial Revolution, to ask him about open platforms. Here's what he said:

In general, the Open Source Hardware innovation model of the DIY Drones dev teams is designed to beat proprietary innovation models in the speed and cost dimensions, but not necessarily in features or performance. We've developed several autopilot systems with an average dev time of one year and a cost of $0 (all volunteer labor). Comparable commercial systems can take 3-5 years and up to several million dollars. Our systems are designed to approach theirs in performance at 1/10th to 1/100th the cost, which is only possible through an open innovation model.
Indeed, I think in some newer domains, such as multicopters, this may go even further. We're already seeing the open source projects starting to pass the proprietary ones in features, if not performance. And as commodity sensor and processor technology gets cheaper and more capable, thanks to smartphones, the amateur designs could pull ahead of the pro ones in performance, too, simply because they can use new technology sooner. This is why DARPA used our model to create UAV Forge, which is designed to let community-based UAV technology compete with traditional aerospace industry technology in the hopes that it will prove superior.

Between the exponential growth of computing power and the work of groups like DIY Drones, the cost barrier to participation in robotics and AI is falling away. When that happens, open source robotics will take off.

And contrary to what some observers have argued, openness can be profitable. Every tech startup in the last 10 years has used open source in some way. Facebook and Apple's iPhone owe much of their success to their open app platforms. Red Hat, one of the leading Linux providers, is an $11 billion company, with revenues of more than $1 billion. The trick for robotics and AI companies will be to figure out the right mix of what's proprietary and what's open.

The Conversation (0)

How the U.S. Army Is Turning Robots Into Team Players

Engineers battle the limits of deep learning for battlefield bots

11 min read
Robot with threads near a fallen branch

RoMan, the Army Research Laboratory's robotic manipulator, considers the best way to grasp and move a tree branch at the Adelphi Laboratory Center, in Maryland.

Evan Ackerman

“I should probably not be standing this close," I think to myself, as the robot slowly approaches a large tree branch on the floor in front of me. It's not the size of the branch that makes me nervous—it's that the robot is operating autonomously, and that while I know what it's supposed to do, I'm not entirely sure what it will do. If everything works the way the roboticists at the U.S. Army Research Laboratory (ARL) in Adelphi, Md., expect, the robot will identify the branch, grasp it, and drag it out of the way. These folks know what they're doing, but I've spent enough time around robots that I take a small step backwards anyway.

This article is part of our special report on AI, “The Great AI Reckoning.”

The robot, named RoMan, for Robotic Manipulator, is about the size of a large lawn mower, with a tracked base that helps it handle most kinds of terrain. At the front, it has a squat torso equipped with cameras and depth sensors, as well as a pair of arms that were harvested from a prototype disaster-response robot originally developed at NASA's Jet Propulsion Laboratory for a DARPA robotics competition. RoMan's job today is roadway clearing, a multistep task that ARL wants the robot to complete as autonomously as possible. Instead of instructing the robot to grasp specific objects in specific ways and move them to specific places, the operators tell RoMan to "go clear a path." It's then up to the robot to make all the decisions necessary to achieve that objective.

Keep Reading ↓ Show less