Does NASA Have a Robotics Strategy? Did It Ever Have One?
Military pushes for emergency robots as skeptics worry about lethal uses, Washington Post
“Most of the [DARPA Robotics Challenge] entrants resemble humans, with two arms and two legs, and could be cast in Hollywood’s next futuristic blockbuster. But there is also “CHIMP,” developed by Carnegie Mellon University, a squat, long-armed machine that uses wheeled treads to get around. Another, named “RoboSimian,” is a four-legged “ape-like” creature developed by NASA that, depending on how its limbs are situated, is also strikingly arachnid-like.”
NASA Centennial Challenges Program: Space Robotics Challenge Request for information, NASA MSFC
“The robots involved as testbeds for the challenge would be both the Valkyrie and Robonaut 2 humanoids, as examples of surface and in-space robots. … The target platforms for this challenge are the Valkyrie and Robonaut 2 robots at NASA JSC. … Valkyrie and Robonaut 2 platforms would be available for preparing for the final challenge. Robots would be accessible at NASA Johnson Space Center, and NASA is currently seeking separate proposals for hosting Valkyrie robots around the country.”
Keith’s note: The next event is 5-6 June in Pomona, CA. There is no mention from NASA or JPL about Robosimian’s participation in this event. JSC’s Valkyrie robot is a no-show since it placed last in 2013 (they are not listed as a qualified team). Is JSC still funding Valkyrie? Is NASA HQ? The Centennial Challenges RFI seemingly speaks of Valkyrie as if it is still a viable ongoing/future program. But it is not clear who pays for it. What is JPL spending on Robosimian? Why is it not part of the Centennial Challenge effort along with Robonaut and Valkyrie? How are Valkyrie and Robosimian related to Robonaut? Does NASA actually have a coherent, focused robotics program or are various centers and directorates just doing whatever they want (because they can/lack of managerial guidance)? NASA spends all this money on these cool things, puts out a few flashy news things, and then they go silent about what they are doing. If you ask for a plan, they have none to offer and they just mumble #JourneyToMars and #YearInSpace.
If there is a strategy here, I do not see it.
Keith’s update: I stand corrected NASA has posted 2015 NASA Technology Roadmaps: Robotics and Autonomous Systems. I am not sure that a “roadmap” is a “strategy” or a formal “plan” – but this is NASA, so words like these are never clearly or consistently defined and are often used interchangeably. To my earlier point: if you search the document for the word “Valkyrie” you will see that this robot is not mentioned. Neither is “Robosimian”. But “Robonaut” is mentioned several times. Not exactly an inclusive document that references all of what NASA is doing – and why.
– NASA JSC’s Valkyrie Robot Tied For Last Place in DARPA Competition, earlier post
– NASA JSC Has Developed A Girl Robot in Secret (Revised With NASA Responses), earlier post
– JPL Rolls Out Robosimian While JSC Hides Valkyrie, earlier post
Robot development is a massive field, and there is no way NASA can cover it all. Most of the needed talent to build and program already resides in universities and small and large companies both long on ideas and short on resourcew. NASA can play a vital role is providing the modest resources needed to enable these brilliant students and entrepreneurs. The annual university robotic mining competition at KSC, for example, has become an effective incentive for aspiring mining engineers from across the country. A NASA SBIR was helpful in developing the Da Vinci surgical robot, and NASA invests in development of specialized robots such as the Mars rovers which need more AI and could have closer ties to the automated vehicle industry. I feel logically NASA should also invest more in robotics for aviation, including autopilots for commercial flight, nonmilitary drones,etc. There isn’t a single national robot, or a single purpose or engineering development needed for robotics, so it is not necessary for NASA to have a single robotics program any more than it is necessary for DOD or the academic community.
The problem is that you have multiple programs inside NASA that work in their own little fiefdom with not a lot of information sharing or coordination with other centers. Programs start and stop based on what funding can be scrounged up from here and there, etc. This is true of areas besides robotics, btw. Having worked on some multi-center NASA R&D before, I can say the turf battles can be intense.
NASA also has the problem that a combination of a static civil service workforce and lousy pay and benefits at the contractors makes it hard to attract and retain staff that have cutting edge skills that are in demand in private industry.
I agree with your points. R&D needs consistent funding so we don’t spend half our time looking for resources. And we need better sharing and collaboration. Some leaders defend their turf. Good leaders have the confidence to invite others onto their turf.
The problem is that all centers, the directorates inside the centers, and so on down all see the others as rivals for their money and power. The CS managers see their importance as being measured by the number of bellybuttons that report to them. (One side effect is that a contractor that figures out how to do the job with fewer headcount won’t be looked on favorably. Seriously – it means the budget will go down next FY, not that you get more money to use elsewhere.
I agree. If we accept the status quo we will accomplish little, so we must change the paradigm. But how do we start?
In other words if you or I (as individuals) can see what NASA is doing – but NASA cannot/ or will not describe it – then there’s not a fundamental problem with NASA?
I agree its a problem, but more with PAO than robotics. Maybe PAO could use a full-time robotics reporter.
If you and I as interested individuals can see it, have access to the reports, and any scientific papers then no, there is no fundamental problem.
It would be unfortunate if NASA went farther into PR than it already does. Very few people participating here like the ‘old space’ business model. Yet excessive central planning, short term focus, overreliance on PR, and marginalization of engineering and engineers is exactly how ‘old space’ operates. The results are stark – it takes longer and costs more to get anything done if anything gets done at all.
Do we want NASA scientists (and scientists in general) fussing about roadmaps, strategic plans, and shared services or do we want them pushing the technical boundaries? I prefer the latter. When there’s something meaningful to report, report it. Technology shouldn’t be the handmaiden of the 24 hour news cycle.
NASA’s Office of the Chief Technologist now has draft 2015 technology roadmaps. Technology Area (TA) 4 is Robotics and Autonomous Systems. TA 4 can be found here: http://www.nasa.gov/offices…
The document does not mention Valkyrie or Robosimian but it does mention Robonaut. Not exactly a comprehensive roadmap.
It gives a pretty good spectrum of technologies for NASA R&D. The part I question is that every piece of work must be stovepiped to the requirements of a unitary “NASA Mission”. NASA’s mission is to use its unique capabilities for both in-house and funded R&D to create practical benefits for America. If we can make a better mining robot, or medical robot, or aerial environmental monitoring robot, then we should do it even if it is not (realistically) needed to explore Mars or Europa.
Back at the creation of NASA, there was an organization at NASA Hq called the Office of Advanced Research and Technology (OART) that was based on the predecessor organization, NACA. In that shop, the standard — which endured under various names into the early 1990s — was to have at NASA Hq technical subject matter experts in diverse fields of interest who had connections (some better, some worse) to the real experts at the NASA Field Centers, universities and companies. (DARPA was formed on the model, and continues to use it with great success even today.)
At NASA, in the 1990s Administrator Goldin dissolved that organizational model. (A complicated and depressing story in those years…) Although the Space Technology Program was re-established as a separate organization at NASA Hq several years ago, the business model of that organization is radically different. There are very good folks there, but the shop and the budget are too small, and it is now organized along budgetary lines — not technical ones. As a result, there is no basis in Washington for a really solid knowledge to judge technical progress being made (much less anticipate future directions), nor is there deep knowledge of the “right” experts to engage outside of the Agency, except occasionally. Of course, the field centers are consulted for efforts like the technology road maps — however each one has its own interests; with no competent “referees” on the field, the result is comprehensive, but lacks focus. There is no ability to establish strategic technical priorities that make sense.
The issue of robotics technology that Keith has raised is not the fault of the space technology team at NASA Hq (as I said, they are trying really hard); however in my view it just the tip of the iceberg.
Best regards…
The first Robot should have been the one designed for dealing with congress. Shoot couldn’t get the picture to show .. an ED 209.
Usually when anyone hears the word robot they mean some sort of humanized thing, doing something a human would otherwise do, especially physically. Mention of rovers almost fits, to the extent they have camera’s or a drill or move, but even there the term fails a bit, as again your not replacing a human per se (the human would not have been there on Mars to be replaced, yet). Sadly enough, in human spaceflight politics, we can’t even get a common understanding that far, like having a goal to have a couple of robots on the ISS, telerobotic say, doing all that any astronaut does, running experiments, performing maintenance, and so on.
Once upon a time we talked about “software agents” (back in the 90’s). We are told today that we have these -in Google, or some setting in my financial app. That’s not what was envisioned or defined back then, and the same bait and switch has happened with robots, to the detriment of the field. Until a NASA strategy also comes with outcomes-things being envisioned, and progress toward that, including operationally, robots will just keep on being everyone’s vague un-coordinated little technology toys and hobbies trapped before a technology valley of death.
Robotics, all of the above – do it! There should be robots on the Moon and Mars doing all sorts of things preparing to mine, process, build, etc. start now so we’re better at it sooner.