The Robots4Us Challenge

robots goes back much much further. When I was a kid, I watched tons of movies with robots in films like Star Wars and The Black Hole. I was a huge fan of Buck Rogers and Balstar Galactica. I mean the good version of Battle Star, you know, the one that was from Night and had that really awesome theme song. Anyway, robots spectored into some of my favorite television series and movies, and that love goes strong today. One of the things I didn't appreciate as a kid but really understand now

is how complicated a field robotics really is. There are enormous challenges designers face when creating a robot. Even a stationary robot that only needs to complete a few precise movements is a complicated and potentially dangerous creation. Robots that move around come with even more challenges, and a robot that can navigate through variable terrain is a huge design challenge.

Then there's the AI side of the equation. Artificial intelligence is a term that gets thrown around a lot, and people often misuse it or focus on a very narrow definition of it. Most of the discussions I hear about AI are really about strong AI. That's the concept of a machine capable of simulating thought on some level. But AI isn't just about getting a computer to think. In fact,

it's not even primarily about that. AI can involve building systems that allow robots to perceive their environ moments, interpret the data they receive, and respond appropriately. It might include optical systems or voice recognition software. It could require natural language processing so that the way you and I speak becomes intelligible to a machine. It could involve plotting a

pathway through a new environment. Now, if you or I were to walk into a room for the first time, we probably find it pretty easy to make our way to a specific point in the room. Even if there were obstacles in the way, we could reason our way around it. Even taking the pathway that requires the least amount of work to get there. For a robot, that's not necessarily the easiest task, particularly in a room that

has changing conditions in it. What I'm getting at is that building robots designed to maneuver through new and potentially changing environments bring with it a host of engineering and programming challenges. It's a wonder that anyone ever tries to even do it. But that brings us to what I wanted to talk about today, the DARPA Robotics Challenge. Now DARPA is the Defense Advanced Research Projects Agency part of

the Department of Defense. Before nineteen seventy two, it was known as ARPA, and it's the agency responsible for the creation of ARPA Net, which is the predecessor to today's Internet. The organization is dedicated to research and development of new technologies. It was founded in nineteen fifty eight, a year after the then Soviet Union launched the beeping satellite Sputnik. President Eisenhower directed U S Secretary of Defense Neil McElroy, to

coordinate a national space program. The Department of Defense formed ARPA in February nineteen fifty eight as a special agency within the Pentagon to maintain technological superiority over potential adversaries. Originally, the agency had three presidential erectives to follow. The first was to get the U S Space program off the

ground figuratively and literally. The second was to come up with the means of protecting the United States from Soviet missile attacks, and the third was to find ways to detect Soviet nuclear tests. The first director of the agency was Roy Johnson, who aggressively pursued the organization's goals. ARPA was and remains independent from the military services now. This gives the organization a huge advantage in that it can explore high risk research and development projects that the military

services can't pursue. The potential payoffs for these projects might be huge, but there may be no guarantee that they'll succeed. The agency has created a culture of thinking outside the box. ARPA was responsible for much of the early work that would eventually send astronauts to space and to the Moon. As the work progressed, ARPA transitioned and shared its research and development with the newly formed National Aeronautics and Space Administration,

or NASA. ARPA itself was affected heavily by this transition, it stopped looking into the space race. Many people were either let go or moved to other departments, and budgets were cut. In the nineteen sixties, the organization focused on other big projects for the Department of Defense. In the process, we made great strides and several scientific fields, including geology, seismology, and radio astronomy. ARPA was identifying specific goals we needed

to meet, which brought to light engineering challenges. This gave people working with and for ARPA the direction they needed to solve tough problems. Often this would create benefits beyond completing a project. Now. DARPA has been involved with robotics for years. I've talked about driverless cars on this show before, like the ones being designed by Google. Many of the cars in development right now owe a lot to DARPA.

Back in two thousand four, DARPA held the DARPA Grand Challenge, which created a competition between teams to build an autonomous car capable of navigating through a course. The big winner of the two thousand four Grand Challenge was nobody. Not a single team managed to complete the course, but rather than become discouraged, the various mechanics engineers, computer scientists and inventors went back to the drawing board to try again. The challenge was held again in two thousand five, and

this time five teams completed the course. In first place was the Stanford Racing team with Stanley their autonomous vehicle. The course included narrow tunnels and sharp turns. On that team with Stanford was Sebastian Thrunn, who was the director of the Stanford Artificial Intelligence Laboratory. Now he has a different job title. He's the project leader on Google's self

driving car project. In two thousand seven, DARPA held another challenge, this time putting autonomous vehicles in a simulated urban environment complete with traffic signs and signals and other traffic Tartan Racing, A team from Carnegie melon University claimed first place. Five other teams successfully completed the course. Maneuvering through traffic and obeying traffic laws added new parameters to the challenge. Cars

couldn't just plot a pathway and follow it. It's hard for me to get across exactly how big a deal designing an autonomous vehicle really is. It has to do things that you and I take for granted, but that in no way shape or form come naturally to a machine, and that's not where the story ends. Now it's time to talk about the DARPA Robotics Challenge, the finals for which will take place in June of this year, two thous and fifteen. But before we get into that, let's

take a quick break to thank our sponsor. You have a smartphone, what about a smart home? Smart Things has created a super easy way to control, automate, and secure every aspect of your home, and you don't have to be a tech genius to install and use it. That's why they were named one of the top ten coolest Gadgets of the Year by Time magazine. That's why they were one of the two thousand fifteen ce S Editors

Award winners. Smart Things instantly turns your normal home into a smart, connected home where you can do magical things just using your smartphone as the interface. You can control lights, locks, thermostats, security with smart Things. It's all connected through a single app that works on iPhone, Android and Windows phones. Personally, I would use this for lots of different things. Here's one example. I can imagine sitting down on my couch

and getting ready to watch a spooky movie. So I use my app to set the lights to the right lighting level to get that good atmosphere. Maybe turn the thermostat down a little bit, make it a little chilly, you know, really get into it so that I can enjoy the experience. And all for the convenience of the phone. I don't have to get up and do any of this manually. Now it has no required monthly fees and kits. Start at smart Things is affordable for everyone, and for

my listeners, it's even more affordable. Get ten percent off any smart Things home security or solution kit and free shipping in the US. Just go to smart things dot com slash tech. The challenge is this build a robot that is capable of entering a disaster area, navigating to a specific location, and activating machinery to prevent the disaster from becoming worse. The inspiration for the challenge came from a real world incident Gil Pratt from DARPA explod eights.

The Darker Robotics Challenge is a program the DARPA is running to develop technology to make our society more resilient to natural and man made disasters. Uh It was inspired directly by the Fukushima disaster in Japan that had a nuclear meltdown and the release of radioactive contamination as part of it, and our inspiration in that disaster was that if robots had been there during the first day or two of that disaster, the release of the radioactive contamination

could have been prevented. In particular, there were three explosions that occurred, and had robots been able to go in there and open up some valves to release the pressure, uh, the disaster would not have been nearly as bad as

it ended up being. Now, of course, we can't change the past, so um what we need to do is to prepare for the future, and the point of the Darker Robotics Challenge is to develop a kind of technology for humans and robots working together at what each one is best at, with the robots working in a contaminated area and human beings working in a safe zone to be able to emergently handle natural and man made disasters

that may occur in the future. Using the Fukushima disaster as a model, DARPA has created a scenario in which robots will encounter situations similar to what was happening in Japan. The challenge isn't just a single task, it's a series of them. We're trying to emulate a real disaster, and even though we don't expect the next disaster to be the same as what happened in Fukushima, we're using that as our exemplar. And so the robots will have to begin by being placed into a vehicle in a safe

area by their human handlers. The robot will then, under human supervision, have to drive a certain distance in order to get to a particular site, which is similar to the reactor site in Japan, but of course ours is

just a mock up. Having gotten to the site, the robot will have to get out of the vehicle and egress on its own, open a door to get into the site, walk into the building, and then travel or some debris, move some debris out of the way, uh and then do tasks that are similar to what needed to be done in Fukushima. The established tasks present big problems for engineers to solve. A robot capable of driving

a car is different from an autonomous vehicle. Teams will have to build robots that have appendages that can work with vehicles we've designed for ourselves. Navigating through debris is another tough problem. The robots will need to be able to test their pathways carefully before committing to them. But then there's the surprise. Robots don't do well with surprises. Typically,

roboticists design their machines for specific purposes. You define the parameters of what your robots should be able to do, and you ignore anything that falls outside those parameters. But surprises, by their very nature can fall outside your normal consideration. So designing a robot that can adapt to situations makes things truly challenging, and in fact, we add some difficulty even beyond a surprise, which is that we make the communications more difficult than it ordinarily would be if there

was no disaster. In real disasters, usually communication infrastructure is degraded and everybody's trying to talk on the radio at the same time, and so we have something called a degraded communication emulator, which is a piece of hardware that makes the communication network between the human beings and the robots much worse than it ordinarily would be if there wasn't any simulation of a disaster going on. And so the teams will have to do well despite a lot

of interruption and a lot of degradation of communication. Now, imagine that you're on one of those engineering teams. Your robot has to be able to drive a vehicle to another location, get out of the vehicle, navigate into a building, climb a flight of stairs, cut a hole in a wall, manipulate cables and controls, and deal with something outside of those events. On top of that, you can't rely on communication systems to guide your robot or keep eyes on

what's happening remotely. It almost sounds like an impossible challenge, but the point is that this pushes roboticists to think through these problems and create new solutions. It forces innovation, and more importantly, it gets us closer to a world in which robots can realistically and reliably respond in emergency situations. Darbas says that we should expect to see robots operating

on the level of a two year old child. They will be able to carry out simple commands autonomously, but we'll need human intervention to chain together a series of simple commands to accomplish complicated goals. This, in turn, will become the foundation for more sophisticated robots in the future, perhaps some with the ability to adapt to new situations and remain flexible and its decision making process. It's an incredibly ambitious challenge and I can't wait to see how

the various teams tackle them. Of course, I'll be reading about the challenge after it happens, but if you are a high school student or you know someone in high school who has an interest in robotics, you may get to witness it firsthand. That's because DARPA is holding the Robots for Us Challenge, which gives a pretty cool opportunity

to students. The Robots for Us Challenge is a contest that we're having for high school aged students from the United States and its territories aimed at beginning to address societal issues with robotics, and in particular, there's been a lot of buzz in the media recently about artificial intelligence and robotics and what it might mean for the future. DARPA believes that it's very important for us as technologists to raise these issues with society and to let society

know that they should start thinking about this. And we think high school aged students are of the right age. They're going to be the generation that's going to be most affected by this, and so uh we would like to see what high school students have to say about these issues. I think this is fantastic. While we rush to innovate in robotics, we also must think about the impact those developments will have, not just the direct impact of this technology, but the social implications that could have

in the future. And while I have some thoughts on the subject, I really do love to hear what younger people have to say about it. They often approach problems with a perspective that simply would never occur to me, and Darba recognizes that potential as well. So how do students get involved? Students can enter the contest by going to the main DARKA Robotics Challenge website, which is the Robotics Challenge dot org. That's just one word, the Robotics

Challenge dot org. And there they'll find a icon for the Robots for Us contest, and you click on that icon and it will take you to the forms in order to register. The submission is in the form of a two to three minute video clip that deals with some issue in robots and society. Now, this is right in my wheelhouse. I do videos like this every week. Sadly, I'm a bit outside the brackets for the contest having graduated high school back in but if I were still

of the appropriate age, who would see my submission? The judges of the video entries are going to be roboticists and ethicist for our first pass, and then a second pass on the winners will be made by Darker program managers and from that we will choose five winners. These don't have to be solo projects. You can get help, but each submission needs one person designated as the representative of the video. That's the person who is eligible to win the prize and the prize that's a trip to

the Robotics Challenge finals. And in addition, Darker will not only pay travel expenses for that student to come to the Darker Robotics Challenge, but also pay the travel expenses for adult chaperon for that student. Let's take another quick break for our sponsor. When you're selling online, getting your orders out the door quickly can be tough. That's why you need ship station dot Com. It's the fast, easy way to manage and ship all your orders all in

one place. With ship station dot com, you can easily import customer orders from eBay, Amazon, Etsy, and more than fifty other popular market places and shopping carts. Then easily create shipping labels for all the top carriers, including UPS and fed X. Plus, get a free USPS account that gives you access to deeply discounted USPS shipping rates, the same rates once reserved for Fortune five hundred companies. It's

no wonder. Ship station dot Com is the number one choice of online sellers with an incredible customer satisfaction rating. Right now, you can try ship station dot com free for third d days, Plus you can get an additional month free only if you use the offer code text stuff. Don't wait. Go to ship station dot com and before you do anything else, click on the microphone at the top of the home page and type in tech stuff. That's ships station dot Com inter text stuff add the challenge.

The students will present their thoughts on the future of robots in our society to the various teams present. They will have a chance to watch the teams compete with one another and complete the tasks. This is the sort of stuff I would have loved to do as a high school student. Heck, I want to do it right now. Submissions are due by April one. You can find more information at DARPA's website for the Robots for Us Challenge at www dot the Robotics Challenge dot org. DARPA has

another challenge objective right now. It's the Cyber Grand Challenge, and as you might guess, it has to do with Internet security. This is becoming increasingly important as more critical systems have connectivity to the network of networks. Take commerce for example. Every year, trillions of dollars of business rely upon the Internet in some fashion as attempting target. And as we add more devices to the Internet in our quest to create the Internet of Things, it's only going

to get more complicated. A connected gadget could serve as a network vulnerability, allowing attackers and entry point into a network. Now, it might be hyperbole to suggest that a refrigerator could lead to the collapse of a bank, but we do need to pay close attention to potential security hazards and DARPA has come up with a creative challenge for programmers. You see, attackers look for soft spots in a network. They just need to find in one vulnerability to exploit.

Their job is relatively easy. Now, designing a secure network that's much more difficult. It's just about impossible to anticipate all potential weak points, and so if you want to make sure your system is secure, you put it to the test on the Capture the Flag tournament circuit. In this circuit, hackers attempt to reverse engineer software in an effort to discover flaws in the programming. Anything that can

be exploited is a potential victory. What DARPA is looking for is an automated system, a computer that can do the same thing as hackers. In an ideal world, a computer program could analyze billions of lines of code, discover any flaws, and even write a patch to remove the vulnerability before anyone could take advantage of it. In two thousand and sixteen, DARPA will hold an all computer Capture the Flag tournament. It will take place during the def

Con conference in Las Vegas. Will computers be able to dissect software with the same precision as human hackers. We'll have to check back in two thousands sixteen to find out. I want to thank Gil Pratt for talking with me about the Robotics Challenge and the Robots for US competition, and I can't wait to hear more about how it all turns out. I'm also curious what the next big challenge from DARPA might be. Got any ideas While we're on the subject, If you have suggestions for future tech

stuff topics, you should throw them my way. My email address is tech stuff at how stuffworks dot com. You can also get in touch with me on Twitter, Facebook, and Tumbler. You can find me with a handle text stuff hs W. Now, I'll talk to you again really soon for more on this and thousands of other topics. Because it has to works dot com. Che