NASA JSC Unveils 'Valkyrie' DRC Robot, IEEE Spectrum
"While NASA's official position is that Valkyrie is a genderless humanoid (as is Robonaut), the robot does have some features that we would call unmistakably female. For example, there's the name of the robot. "Valkyrie" (the roboticists call it "Val" for short) refers to the female figures in Norse mythology who decided which warriors fought valiantly enough in battle to be taken to Valhalla when they died."
Keith's update: This placeholder website just appeared. But it only shows Valkyrie's back - not her front. There was a Facebook page up until yesterday that referred to Valkyrie as a female and said that Robonaut was her "brother". NASA JSC was originally directed by NASA HQ to make this robot gender neutral. JSC then ignored that direction and deliberately made the robot overtly female - at least its upper torso. Now they deny that it is female or that it was ever intendeded to evoke or portray a female characteristics in any way. This project was also kept secret from a large number of JSC engineers who would otherwise be aware of such activities. Someone is not being truthful.
Just for the record folks, I have no problem whatsoever in making robots and other NASA hardware more approachable to people in ways that make them feel included. None whatsoever. And if you are going to make a space robot why not make it look cool? Lets see more of it! What is troubling is how NASA JSC played favorites with one specific media outlet to the exclusion of all others - and still does; how NASA JSC is being less than honest when they say that there was no intent to imply gender identity on this robot when in fact multiple well-placed sources state that this was intentional from the onset at JSC; and that HQ PAO is now trying to tell me that a robot with an overtly female chest configuration and a female name is not trying to invoke a female identity. I'm not that stupid - neither is everyone else.
Why can't they just admit the obvious to the public in the same way that they discuss it internally? Its time for NASA JSC to be honest and stop trying to weasel out of the obvious by hiding behind HQ PAO responses. Ellen Ochoa knows the whole story.
Keith's note: I submitted the following questions to NASA PAO, JSC PAO, and Mike Gazarik - the reply from NASA HQ is listed below:
1. why has there been no PAO material issued about this robot? Why is the only information online at IEEE? Why were other media not afforded an opportunity to interview people working on this program? NASA chose not to promote R5 in the press prior to the DARPA Robotics Challenge both because it is part of a competition and because of the extraordinary workload the team has faced.
NASA: In many challenges, teams wait to reveal their designs until just before the competition. Also, the NASA team lost two weeks of work when JSC was closed due to the government shutdown. They've been very busy working to get R5 ready for next week.
IEEE approached NASA JSC and the R5 team last summer and requested access to R5 and interviews with the team at JSC. In return for access to the team and the robot as it was being built, the IEEE reporter agreed to hold his story until now to avoid disclosing the NASA design prior to the robot's completion. The reporter's research was conducted last October, before NASA formally named the robot "R5."
The robot will be competing in the DARPA Robotics Challenge http://go.usa.gov/We9W Dec. 20 and 21 in Homestead, Fla. NASA will participate in press events at the challenge -- journalists will be able to interview team members and see the robot at the competition.
R5 was referenced in a NASA press release we issued last Monday about R2 getting legs http://go.usa.gov/ZcPz . In that release we mention R5 and link to the DARPA Robotics Challenge website.
NASAWatch Follow up: Your answer makes no sense. First you say that you don't want to reveal anything and yet you let IEEE all over the place - with an exclusive - allowing them to take photos and have preferential access to government personnel - and then you refuse to allow any other media to have similar access after that story is released. You don't even post any imagery online even though the secret of its appearance is now all over the place. You just said that the competition is on 20 Dec and yet IEEE has revealed the design "prior to the robot's competition". Did IEEE sign an NDA on this? Can you point me to the NPD or other agency policies whereby NASA allows exclusive access to something that (as you note below) contains restricted technology? Was IEEE limited in what it could describe or photograph?
NASA Follow-up response: Keith, as the recipient of numerous NASA "exclusives" over the years, including private budget briefings and policy discussions with previous NASA Administrators, I'm amused by your feigned outrage over our unintentional exclusive to IEEE Spectrum on R5.
IEEE was the first to ask for an insider's video view of the robot as it was being built. Their request came in before the government shutdown and before the need for the team to accelerate work to due to undiscovered technical hurdles.
Due to the government shutdown, after IEEE's visit the team did not have the time to accommodate additional media visits and still meet their already extremely aggressive schedule. Accommodating such visits would have required the team to go into downtime and interfere with the lab environment.
IEEE did not sign an NDA. Their access was not originally planned to be exclusive. Events warranted flexibility in our communications approach to stay on schedule. As a very kind professional courtesy to NASA, IEEE held the story to allow us to make our initial announcement and point people to the competition for media avails and interviews. Journalists will have access to our team and robot at the competition Dec. 20-21 - I encourage you to register now and come on down.
2. how much did this robot cost? Who paid for this robot?
NASA: NASA received $3 million from DARPA and matched that with $3 million from NASA. We also received $1.5 million from the State of Texas Emerging Technology Fund and $100,000 from Houston's Jacobs Engineering. These funds were mainly used for design and development of the first robot.
3. What specific NASA program(s) sponsored this robot?
NASA: NASA's Space Technology Mission Directorate's Game Changing Development Program funds robotic research at the agency's Johnson Space Center in Houston and provided NASA funding for this project.
4. Who actually built this robot - was this done in house? Were contractors used - if so which one(s)? How were their services procured?
NASA: An R5 team of 55 people includes NASA engineers from the agency's Johnson Space Center in Houston working with engineers from Houston's Jacobs Engineering and Oceaneering Space Systems, as well as students from the University of Texas at Austin, Texas A&M University and the State University of New York at Buffallo.
5. Will there be more than one robot constructed?
NASA: NASA does not have plans to build a second R5, but hopes that other government, university or commercial partners may choose to build upon the technology developed for R5 for applications that can aid in disaster recovery or mitigation efforts, or in commercial applications here on Earth.
6. Will the plans for this robot be made available to the public in open source format?
NASA: Because the team consists of partners outside of NASA who have invested time and financial resources into the project, some parts of the IP for R5 may be restricted.
NASAWatch Follow up: Who is on the team? How did they partner with NASA - SAAs? MOUs? Contracts? What aspects of this robot are specifically is restricted - and why? Who actually owns the design of this robot? Is anything on this robot ITAR restricted?
NASA Follow-up response: NASA and NASA contractors - contracts - CRAVE II, JETS, EPIC, Bioastronautics), UT, A&M, SUNY, IHMC Agreements - Cooperative Agreements, then ETF funding from the State of Texas for universities. We're not allowing photos "under the skin," as NASA is still early in the process of disclosing and filing for protection of intellectual property. This is the same approach we have taken on other robots, like Robonaut 2.
7. Why was this robot overtly designed with a specific gender? Was this a program requirement?
NASA: R5 is neither male nor female, it is a humanoid robot. While humanoids do not have a defined gender, they have certain characteristics that are found in human beings. NASA hopes to show that robotic designs do not have to be of a specific gender-design. R5 was designed with form and function in mind, with consideration to the challenges the upcoming competition poses.
NASAWatch Follow up: Oh please, everyone calls it by a girl's name and says "she" when referring to it. The robot has a blatantly female upper torso. Why anyone would intentionally put that obstructive feature on the front of the torso so as to restrict the movement of its arms is baffling. Bad ergonomics. If it is a design feature with an intentional purpose, can you tell me what specific subsystems are inside the upper portion of the torso such that the obvious pronounced projection (female shape) is required - a feature, by the way, that Robonaut does not seem to require?
Photos from IEEE
NASA Follow-up response:The short answer is - it made the most functional sense. A longer answer: The engineering analysis on the form was done before the idea to model the robot's form - the form was driven solely by function and engineering need. R5 is completely self-contained, that is to say it has all its computers, power source and cooling onboard the robot. To keep the center of mass as close to the geometric center of the robot, we had to push the battery as far forward into the torso as we could. The battery weighs more than 30lbs. If we had not, and instead chosen an R2 style backpack, we would have had a very 'hunched over' walking style and therefore it would have been very challenging from a manipulation standpoint. In addition, we also needed to have an articulating waist so that the manipulation workspace of the robot could be repositioned while standing and not moving. The linear actuators for the waist then were positioned around the battery and subsequent power converters. Given the kinematics of the waist and therefore subsequent angle of the actuators, it led to a feminine humanoid feature. Currently, R2 only rotates at the waist - but does not 'bend.' What we are talking about is the functional equivalent of the lumbar portion of a human spine. Robonaut does not have that and therefore packing strategies between the two robots were very different. A spine is very necessary in center of mass management when walking in a gravity field, something Robonaut doesn't have to contend with.8. The total possible prize to be awarded by DARPA Robotics Challenge is $34 million. If NASA enters the Valkyrie robot and it wins where will the prize money go?
NASA: As the owner of the competition, you should check with DARPA concerning the expected value of their final prize purse and their prize award criteria.
The first round of trials for the competition takes place later this month. The final prize money will not be awarded until there is a winner during the final competition in December 2014. NASA is seeking guidance from our General Counsel on whether the agency or team could accept any prize money if we were to win the competition a year from now.
NASAWatch Follow up: Do any of the non-NASA partners share in the prize money? How is their percentage determined?
NASA Follow-up response:We're waiting for legal opinion on what we would legally be allowed to do if we won money from the competition and how any prize monies might be managed. We have plenty of time for gaining a legal opinion about prize money -- the team is focused on the trials taking place next week.
9. Is the Valkyrie program related to Robonaut? If so, how? If not, why not?
NASA: Yes. R5's physical design comes from prior technology developed for Robonaut, the X1 Exoskelton and several other robot spinoffs. R5 has Robonaut's arms and neck, the X1's legs and shoes, a hand derived from the RoboGlove, the battery from NASA's Chariot rover and the navigation sensors from NASA's Centaur rover.
Additional Follow up questions
1. NASAWAtch: Did NASA HQ direct NASA JSC to make the robot gender neutral? Did JSC agree to that HQ direction? Did JSC project staff then ignore direction from NASA HQ or did JSC management not send HQ direction down to JC project staff?
NASA: NASA's policy -- longstanding -- is that all robots are gender neutral. They are machines. The R5 team at JSC is focused on preparing for next week's DARPA Robotics Challenge. Our communications coordination is a collaborative effort between interested offices at HQ and JSC.
2. NASAWAtch: There was a Facebook page online by one of the team members at https://www.facebook.com/nasa.valkyrie?fref=ts that what since been removed. It referred to Valkyrie as female and that Robonaut as her brother. Why was that page taken offline?
NASA: The Facebook page was publicly posted before appropriate agency reviews were made and was taken down. We review public web and social media sites for clarity and accuracy.
3. NASAWATCH: Multiple JSC sources report that this project was kept secret from other JSC personnel. What level of secrecy/proprietary information restriction was applied to this project and what is the contractual basis for such secrecy/proprietary information restrictions?
NASA: It's not uncommon for teams that are preparing to compete in challenges to prefer privacy as they build their entries. JSC management was fully aware of this activity and it was funded by NASA's Space Technology Mission Directorate at HQ. It also was no secret that the team had entered the competition.
4. NASAWATCH: Did any foreign nationals work on this project? If so what clearances were required for them to work on it?
NASA: No foreign nationals worked on R5.