LEGGED SQUAD SUPPORT SYSTEM (LS3): DARPA's Tactical Technology Office and the U.S. Marine Corps awarded Boston Dynamics a 30-month, $32-million contract to deliver a legged robot to support troops.
Image: COURTESY OF BOSTON DYNAMICS © 2009
-
The Best Science Writing Online 2012
Showcasing more than fifty of the most provocative, original, and significant online essays from 2011, The Best Science Writing Online 2012 will change the way...
Read More »
Within the next three years, the U.S. military will test the feasibility of sending a quadruped robot out into the field as a trusty pack mule to carry supplies for its troops, wherever they go. If the testing goes well for Boston Dynamics's Legged Squad Support System (LS3), company founder Marc Raibert will have come a long way from the one-legged hopping robots he pioneered in the 1980s.
Actually Raibert has already come a long way, to the point where the U.S. Defense Advanced Research Projects Agency's (DARPA) Tactical Technology Office and the U.S. Marine Corps awarded his company a 30-month, $32-million contract last week to deliver a prototype LS3. This would be the first step in fulfilling the military's call for an autonomous, legged robot that can carry up to 181 kilograms of supplies for at least 32 kilometers without refueling.
The military already uses unmanned aerial vehicles for reconnaissance or to attack enemy targets, and DARPA has sponsored several contests in recent years to determine the feasibility of developing autonomic ground transportation. Automation has been much more difficult to introduce to the infantry, however, because of the need to traverse rough terrain where robots operating on wheels or tracks cannot go.
The LS3 is the latest in a series of legged robots developed by Raibert, who got his start in 1980, when he founded a robotics workshop at Carnegie Mellon University in Pittsburgh called the Leg Laboratory. Raibert moved the Leg Lab to Massachusetts Institute of Technology in 1986 after becoming a professor of electrical engineering and computer science there. The lab, which he directed until 1995, did good work, he says, "but it always felt like we were kicking the can down the street, making scientific progress, but not worrying about a system that could operate out in the real world."
Boston Dynamics, which Raibert founded in 1992, scored a breakthrough in 2003 when DARPA began funding the development of BigDog, a 75-kilogram mechanical workhorse and the LS3's predecessor. BigDog did not use cameras or laser sensors to determine its location. Instead, it stepped first and then reacted to the terrain, quickly determining its position at any given time and comparing that with its desired position, immediately taking corrective action based on the difference between these two.
But BigDog was not an instant success. "In the first year or so it was creaky, I'll tell you," Raibert says. Getting the robot's legs, each of which featured four joints, to move in a synchronized fashion—not to mention adjust dynamically to slippery surfaces such as ice—proved challenging. Over time, however, the quadruped robot's endurance and stamina improved—as did its software for balance and control—to the point where it could maneuver out in the real world. "There reached a point about three years ago with BigDog where we started to see the light at the end of the tunnel," he says.
With BigDog, Boston Dynamics demonstrated to DARPA the feasibility of a legged robot, which was significant for its potential to traverse sand, rocks, mud and snow under conditions that would strand a robot that ran on wheels or tracks. The company tested its technology at Marine Corps Base Quantico in Virginia, where it showed that BigDog could carry about half its own body weight up difficult hiking trails and that it could carry 1.5 times its body weight on flat level terrain. Testing also showed that it could operate without a driver, using an onboard computer vision system to follow a leader who wore a special vest (not unlike an actual dog tagging along behind its master), and that it could travel autonomously using a global positioning system to locations specified on a map (correcting for balance on uneven terrain as it went). The BigDog could march 19 kilometers before needing to refuel.
The LS3—which could end up resembling the Imperial walkers from the Star Wars movies, albeit on a smaller scale—is essentially "BigDog on steroids," Raibert says. The LS3 will have a leg up on its predecessor via its ability to travel autonomously without the need for preset coordinates or soldiers to wear special clothing. The new quadruped will have "a higher level of maturity in terms of system and autonomous operation," Raibert says. "The challenge is to make this fit within a certain size." One of DARPA's specifications is that the LS3 weigh no more than 570 kilograms when fully loaded.
Although LS3's propulsion system is still being designed, BigDog's used a two-stroke, go-kart gasoline combustion engine to drive a hydraulic pump. Whereas BigDog had an 18-horsepower engine, Raibert says he wants the LS3's power plant to be at least 40 horsepower. The LS3 will also have to operate no louder than 70 decibels, significantly quieter than BigDog. (This video of BigDog released by Boston Dynamics in 2008 demonstrates just how loud BigDog can be when on the move, thanks at least in part to its engine.)
The first 30 months of the project constitute phase 1, at the end of which Boston Dynamics will have to deliver two prototype LS3s that can carry the required weight (181 kilograms) a required distance (at least 32 kilometers) across a relatively flat surface. The LS3s will also have to be able to run up to 16 kilometers per hour and feature at least a rudimentary version of the systems it will need to operate autonomously.
Boston Dynamics has enlisted the aid of engineers and scientists at defense contractor AAI Corp., aircraft-maker Bell Helicopter, Carnegie Mellon's National Robotics Engineering Center, NASA's Jet Propulsion Laboratory and motion-control systems provider Woodward HRT to develop the LS3's hydraulics, propulsion and guidance systems. Although JPL previously did some work with Boston Dynamics on the BigDog's vision guidance systems, this is the first time the company has collaborated outside its walls to develop its robots.
See what we're tweeting about
36 Comments
Add CommentWhy not use mules? Our lust for technological replacements using oil and not sand friendly seem misplaced given our current wars geography.
Reply | Report Abuse | Link to thisusing mules would be non-PC and I'm sure PETA would sue for cruelty to animals
Reply | Report Abuse | Link to thisThat's a good question actually. I suppose the problem with mules would be that you have to feed and water them even when they're not being used (and perhaps even during use they require more food & water than the robot needs fuel).
Reply | Report Abuse | Link to thisBut there certainly are many advantages to mules and I hope they considered them seriously before deciding to fund this robot. For example mules are quieter and they're definitely sure-footed. Plus I'm sure you could buy a lot of mules for the price of one of these robots. Lastly they could be eaten in an emergency ;).
That's a good question actually. I suppose the problem with mules would be that you have to feed and water them even when they're not being used (and perhaps even during use they require more food & water than the robot needs fuel).
Reply | Report Abuse | Link to thisBut there certainly are many advantages to mules and I hope they considered them seriously before deciding to fund this robot. For example mules are quieter and they're definitely sure-footed. Plus I'm sure you could buy a lot of mules for the price of one of these robots. Lastly they could be eaten in an emergency ;).
You know they use mules to help fight fires here in Northern California. They can go where no motorized vehicle can go and pull bladed equipment to cut fire breaks and other dangerous tasks on nearly perpendicular mountain sides. They are wonderful animals.
Reply | Report Abuse | Link to thisThe Marine Corps has and does use mules for mountain applications. However, the whole 'care and feeding' equates to more logistics and can complicate things if the mule becomes sick, tired, etc. Plus a robot does not have come with the disposition - remember the saying "Subborn as a mule"?
Reply | Report Abuse | Link to thisThat's what's wrong with the American military, spending hundreds of millions of dollars for something that can be done for pennies.
Reply | Report Abuse | Link to thisWhy don't they just do what the North Vietnamese used to do and hire porters for ten cents a day? That's how the NVA were able to run circles around US infantry who were weighed down with up to a hundred pounds of gear strapped to their back in 100 degree weather.
Sorry for the doublepost above.
Reply | Report Abuse | Link to thisRegarding the stubbornness of mules: mules can be stubborn but robots break down and need parts. Either way your stuff isn't moving.
And as far as porters are concerned I wonder if the military would prefer not to reveal the details of their movements to natives in many circumstances.
If necessary they can be put on leashes. (I'm not joking... Well, maybe a little)
Reply | Report Abuse | Link to thisHas anyone forgotten that mules die? These people are in danger zones. What is supposed to happen if 3 of their mules die with full loads? Do you realize that time wasted and mishaps cause the loss of lives out there?
Reply | Report Abuse | Link to thisI remember seeing a prototype of this type of vehicle in a Time-Life book back in the 1960's. It was due to be put into use in a couple of years. Don't hold your breath.
Reply | Report Abuse | Link to this"What is supposed to happen if 3 of their mules die..."
Reply | Report Abuse | Link to thisAt least machines never break down?
Hiring porters would be considered contract work much like Haliburton and other companies presently offer, which in case you hadn't notice everybody seems to think that is wrong even though every war since Americas beggining has involved such contract work.
Reply | Report Abuse | Link to thisA mule cannot autonomously go to a directed location. A mule does not always go where it is supposed to if you're beating it with a 2 x 4. It's apparent that this is a needed technology if you don't look at it through agenda driven eyes.
Reply | Report Abuse | Link to thisWhy would you take animals into a human combat zone? We don't need to do that anymore -- we've moved on. Other than fuel, you don't have to feed, water, rest, and pet a robot. You also don't leave a trail of crap behind, or worse, have to pick it up as you go along. If you're such a technophobe, you're on the wrong website.
Reply | Report Abuse | Link to thisA similar version of this technology would be nice to land on other worlds, nuke-powered and bristling with sensors and cameras.
Reply | Report Abuse | Link to thisI am friends with two mules. They are not always good workers and can be very moody. AND these guys will bite. Sometimes, they side kick if you annoy them...They have good days and bad days...just like us.
Reply | Report Abuse | Link to thisAfter the cuts at NASA - that's exactly who will be visiting other worlds for a foreseeable future. You make a great point.
Reply | Report Abuse | Link to thisUse a bicycle, very good for going downhill.
Reply | Report Abuse | Link to thisFunny people,
Reply | Report Abuse | Link to thisFor the same reason we do not use Horses anymore (you know we use those Jeep things??!!)
Any animal in the battle field is more of a liability then a machine.
lamorpa,..i remember that too...still waiting.
Reply | Report Abuse | Link to thisIt seems to me some thing more insect like would be better at what they want it to do(more legs)..spider/cockroach
Brilliant Wayne Williamson,
Reply | Report Abuse | Link to thisExcept think terrain, think obstacles, load capability, think mode of transportation... how about centipede, hydraulics or pneumatic with servo motors, 25ft long 4 passenger/supplies with synchronous legs that mimic it's forward counterpart... on flat terrain up to 25mph... Off terrain an agile and adept contortionist with powerful adaptable feet... can scale 10ft vertical over hang as it can contort it's frame up and reach, grab, spike itself to the ground anchoring itself to pull itself up... inch worm like. Armed with an advanced anchor cable launching system uses up to 250ft of retractable spider steel cable with anchors suited to anchor able obstacles... it's last 4 pairs of legs are anchors that holds the other end of the 250 ft cable are retrievable once obstacles such as crossing ravines, 250ft vertical climbs... the first 4 legs detach and using the cable pull themselves at 30mph to latch themselves forming a taunt cable car wire that passengers and cargo can travel 250ft to the other side up to 30mph to reattach with the front 4 legs then retrieve the 4 back legs and 10ft tail that transforms to a kite shape then lofting and retrieval using high speed motors. Between 2 sets of legs (4) will have mini command post and soldier compartment housed in ultra light super strong armor as is it's entire composition with 3 sets of legs these babys are armed to the teeth. Can launch ordinance, missiles, probes (finding the path made simpler and guidance for weaponry), and bullets capable of penetrating armor or targets 2 miles away. 3000lbs payloads. Use NASA engineering with military funding and weaponry build 8 of these per platoon so 36 soldiers could travel across terrains bring supplies and be quite a force on its own launching multiple munitions, advanced target acquisition like aircraft and capable of getting out of Dodge if the going gets tough. Drone helicopters could be programmed to locate it's mate then whisk it back for preparations for their next infiltration or lift fresh supplies to troops, or act as transport to remote destinations. And you will have a mule with a kick, that doesn't mind to fly. Ya gotta love it. Think big.
Quite- when we tell one of these things to make no noies with out nosie they do.
Reply | Report Abuse | Link to thismules make noise when they die.
Reply | Report Abuse | Link to thisImperial Walkers -- Star Wars
Reply | Report Abuse | Link to thisMade a very satisfying "thud" when they hit the ground.
Million$ eh?
No body has said anything about the fear factor,robots don't run away in a fire fight,they go where they are told to with out question,including being used as a shield,and weapons platform.It could be used to help carry the wounded out of harms way,leaving two more men in the fight.I have no doubt there will be a group of smaller bots along used to scout and for other tasks.As for attachment ,I fear that soldiers might risk their lives trying to save a damaged robot.It should be encouraged that the robots not be given names,but that might be hard considering the closed nit group of a combat unit,and if the robot has saved a number of lives.I remember reading somewhere about a soldier in Iraq lovingly picking up the pieces of a bomb disposal robot.Its a whole new world.
Reply | Report Abuse | Link to thisI don't think we can avoid using robots/technology in war or any other event. Let's face it, if we don't use them other countries will despite. In addition, have we not figured out it is our nature to progress in all areas? The ideal world (for me) would have no war, untill we have reached that evelutionary stage of harmony we must move forward. Aside from aiding our military, this technolgy will help amputees, people in wheel chairs. I am sure there are more benifits that can come from investing in this technology than we can now imagine.
Reply | Report Abuse | Link to thisWow, I thought the same thing ....about soldiers getting attached to the robots after I viewed the video. I was surprised at how much it appeared to me like a animal especially while it was slipping on the ice, i actually felt a little sorry for it. I definatly agree with everything you shared. It IS a whole new world. Every thing we have dreamed up has already been made and incorporated into our daily lives or we are already working on it.
Reply | Report Abuse | Link to thisThis if it gets put into the field will be 1st gen, more will follow, doing a better job for less fuel. As much as mules have their pros over things like this, evolution cannot keep up with the onslaught of technology. All things living in the field of war other than our brains will become obsolete to technology. This robotic mule is worth it for the simple fact that you need field time to tweak the next generation which over time will greatly surpass the mule.
Reply | Report Abuse | Link to thisIt seems that this urge to robotics is a part of an Intelligent Design.
Reply | Report Abuse | Link to thisLet's consider the following game:
There is a civilization in disposition of a player. A player can
interfere in history (manipulate polititians and mood of mobs).
The goal is a fastest possible technological development.
Problem 1. A world consists of isolated communities. One part of it (Europe)
makes major advancements in technologies an firearms. How to spread
this knoledge across the world?
Answer - wars and colonization.
Problem 2. The several next technologies (air-space, nuclear, first computers) require a big investment.
How to compel the large multitude of people to pay their money for this?
Answer - make them fear. First through the real big world war and tragedies. Afterwards a cold war will suffice.
Problem 3. A world is divided by iron wall. The development of personal computers requires largest market.
How to unite the world? The answer is obvious.
Problem 4. Robotics require much time and efforts to evolve. A private companies can't afford development because the first robots are too awkward and hard to sell. How the player can organize the money for development?
Answer - again fear. But this time the nature of this fear (terrorism) changes so,
that the old approache (nuclear wepons) is of no use.
There are plenty of military applications for almost everything we have designed. The best designs are a deterrent to physical aggression. When consequences of economical dogma lead to yet another economic recession, it is time again to observe more real stuff coming out from the labs. Now if we can build better and smaller brains than those found in hexapod insects and apply it to nanotechnology. Scary? Maybe not for our children.
Reply | Report Abuse | Link to thisYes, why not? Like when the first automobile couldn’t really compete against horses?
Reply | Report Abuse | Link to thisyes, mules die, and complex technology breaks down, and costs a fortune in time and money to repair.If your mule is dead, get another mule! The military has become obsessed with solving hundred dollar problems with million dollar technology.This is possibly one of the most stupid ideas I have ever heard.Of course, it is a bit sci-fi, so it will appeal to the technical, who probably are scared of mules, and it is probably going to provide a research team and a high tech manufacturer with an endless supply of taxpayers money before it is eventually screpped as impractical or unreliable. If you need mules, use mules!
Reply | Report Abuse | Link to thisOh ye of little brains! Obviously transport bots are only a small step in the journey to fully automated warfare where no humans are involved on our side. By using advanced technology we have reduced casualty rates significantly. Compare casualty rate between Vietnam and Iraq/Afghanistan.
Reply | Report Abuse | Link to thisYou can snivvel all you want but no one cares. You didn't volunteer to go to a war zone and be a human pack mule so piss off you losers.
If I was in a war zone I would prefer a big f#@king robot next to me than a mule. I think the mule angle only works for nerds who have never handled any weapons in their life. The psychological benefits of having a big piece of hardware far outweighs a mule.
Reply | Report Abuse | Link to thisOnce again people this is a science mag, it's about new technologies, not old ideas, if you don't like it read something else.
As a tax payer, I am very concerned about spending millions of dollars on technology that lacks the common sense approach......I cannot believe these scientists are currently using a 2 stroke go-kart engine.....do you have any idea how loud that is? With this approach....how long and how many more millions of dollars will it take these scientists to curb the noise. The enemy will always know where the robots are. I vote for mules....they have been moving cargo for centuries and they are quite and very smart and can go anywhere....Grand Caynon....anywhere. Let's get smart.
Reply | Report Abuse | Link to this