Category Archives: robot rescue

Slow robots and slippery rhetorics

DRC

The recently concluded DARPA Robotics Challenge (DRC), held this past week at a NASCAR racetrack near Homestead, Florida, seems to have had a refreshingly sobering effect on the media coverage of advances in robotics.  A field of sixteen competitors, the victors of earlier trials (it was to be seventeen, but ‘travel issues’ prevented the Chinese team from participating), the teams represented the state of the art internationally in the development of mobile, and more specifically ‘legged’ robots.  The majority of the teams worked with machines configured as upright, bipedal humanoids, while two figured their robots as primates (Robosimian and CHIMP), and one as a non-anthropomorphised ‘hexapod’. The Challenge staged a real-time, public demonstration of the state of the art; one which, it seems, proved disillusioning to many who witnessed it.  For all but the most technically knowledgeable in the audience, the actual engineering achievements were hard to appreciate.  More clearly evident was the slowness and clumsiness of the robots, and their vulnerability to failure at what to human contenders would have proven quite unremarkable tasks.  A photo gallery titled Robots to the Rescue, Slowly is indicative, and the BBC titles its coverage of the Challenge Robot competition reveals rise of the machines not imminent.

Reporter Zachary Fagenson sets the scene with a representative moment in the competition:

As a squat, red and black robot nicknamed CHIMP gingerly pushed open a spring-loaded door a gust of wind swooped down onto the track at the Homestead-Miami Speedway and slammed the door shut, eliciting a collective sigh of disappointment from the audience.

In the BBC’s video coverage of the event, Dennis Hong, Director of the Virginia Tech Robotics Lab, tells the interviewer: “When many people think about robots, they watch so many science fiction movies, they think that robots can run and do all the things that humans can do.  From this competition you’ll actually see that that is not the truth. The robots will fall, it’s gonna be really, really slow…” and DARPA Director Arati Prabhaker concurs “I think that robotics is an area where our imaginations have run way ahead of where the technology actually is, and this challenge is not about science fiction it’s about science fact.”  While many aspects of the competition would challenge the separateness of fiction and fact (not least the investment of its funders and competitors in figuring robots as humanoids), this is nonetheless a difference that matters.

These cautionary messages are contradicted, however, in a whip-lash inducing moment at the close of the BBC clip, when Boston Dynamics Project Manager Joe Bondaryk makes the canonical analogy between the trials and the Wright brothers’ first flight, reassuring us that “If all this keeps going, then we can imagine having robots by 2015 that will, you know, that will help our firefighters, help our policemen to do their jobs” (just one year after next year’s Finals, and a short time frame even compared to the remarkable history of flight).

The winning team, University of Tokyo’s spin-out company Schaft (recently acquired by Google), attributes their differentiating edge in the competition to a new high-voltage liquid-cooled motor technology making use of a capacitor rather than a battery for power, which the engineers say lets the robot’s arms move and pivot at higher speeds than would otherwise be possible. Second and fourth place went to teams that had adopted the Boston Dynamics (another recent Google acquisition) Atlas robot as their hardware platform, the Florida Institute for Human and Machine Cognition (IHMC) and MIT teams respectively. (With much fanfare, DARPA funded the delivery of Atlas robots to a number of the contenders earlier this year.)  Third place went to Carnegie Mellon University’s ‘CHIMP,’ while one of the least successful entrants, scoring zero points, was NASA’s ‘Valkyrie’, described in media reports as the only gendered robot in the group (as signaled  by its white plastic vinyl body and suggestive bulges in the ‘chest’ area).  Asked about the logic of Valkyrie’s form factor, Christopher McQuin, Nasa’s chief engineer for hardware development, offered: “The goal is to make it comfortable for people to work with and to touch.”  (To adequately read this comment, and Valkyrie’s identification as gendered against the ‘neutrality’ of the other competitors, would require its own post.)  The eight teams with the highest scores are eligible to apply for up to $1-million in funding to prepare for the final round of the Challenge in late 2014, where a winner will take a $2-million prize.

An article on the Challenge in the MIT Technology Review  by journalist Will Knight includes the sidebar: ‘Why it Matters: If they can become nimbler, more dexterous, and safer, robots could transform the way we work and live.’  Knight thereby implies that we should care about robots, their actual clumsiness and unwieldiness notwithstanding, because if they were like us, they could transform our lives.  The invocation of the way we live here echos the orientation of the Challenge overall, away from robots  as weapons – as instruments of death – and towards the figure of the first responder as the preserver of life.  Despite its sponsorship by the Defense Advanced Research Projects Agency (DARPA), the agency charged with developing new technology for the military, the Challenge is framed not in terms of military R&D, but as an exercise in the development of ‘rescue robots‘.

More specifically, DARPA statements, as well as media reports, position the Challenge itself, along with the eight tasks assigned to the robotics teams (e.g. walking over rubble, clearing debris, punching a hole in a drywall, turning a valve, attaching a fire hose, climbing a ladder), as a response to the disastrous melt down of the Fukushima-Daiichi nuclear power plant.  (For a challenge to this logic see Maggie Mort’s comment to my earlier post ‘will we be rescued?’)  While this begs the question of how robots would be hardened against the effects of nuclear radiation and at what cost (the robots competing in the challenge already costing up to several million dollars each), Knight suggests that if robots can be developed that are capable of taking on these tasks, “they could also be useful for much more than just rescue missions.”  Knight observes that the robot of the winning team “is the culmination of many years of research in Japan, inspired in large part by concerns over the country’s rapidly aging population,” a proposition affirmed by DARPA Program Manager Gill Pratt who “believes that home help is the big business opportunity [for] humanoid robots.”  Just what the connection might be between these pieces of heavy machinery and care at home is left to our imaginations, but quite remarkably Pratt further suggests “that the challenges faced by the robots involved in the DARPA event are quite similar to those that would be faced in hospitals and nursing homes.”

In an article by Pratt published early in December in the Bulletin of the Atomic Scientists titled Robot to the Rescue, we catch a further glimpse of what the ‘more than rescue’ applications for the Challenge robots might be.  Pratt’s aspirations for the DARPA Robotics Challenge invoke the familiar (though highly misleading) analogy between the robot and the developing human: “by the time of the DRC Finals, DARPA hopes the competing robots will demonstrate the mobility and dexterity competence of a 2-year-old child, in particular the ability to execute autonomous, short tasks such as ‘clear out the debris in front of you’ or ‘close the valve,’ regardless of outdoor lighting conditions and other variations.”  I would challenge this comparison on the basis that it underestimates the level of the 2 year old child’s competencies, but I suspect that many parents of 2 year olds might question its aptness on other grounds as well.

Having set out the motivation and conditions of the Challenge, in a section titled ‘Don’t be scared of the robot’ Pratt  turns to the “broad moral, ethical, and societal questions” that it raises, noting that “although the DRC will not develop lethal or fully autonomous systems, some of the technology being developed in the competition may eventually be used in such systems.”  He continues:

society is now wrestling with moral and ethical issues raised by remotely operated unmanned aerial vehicles that enable reconnaissance and projection of lethal force from a great distance … the tempo of modern warfare is escalating, generating a need for systems that can respond faster than human reflexes. The Defense Department has considered the most responsible way to develop autonomous technology, issuing a directive in November 2012 that carefully regulates the way robotic autonomy is developed and used in weapons systems. Even though DRC robots may look like those in the movies that are both lethal and autonomous, in fact they are neither.

The slippery slope of automation and autonomy in military systems, and the U.S. Defense Department’s ambiguous assurances about their commitment to the continued role of humans in targeting and killing, are the topic of ongoing debate and a growing campaign to ban lethal autonomous weapons (See ICRAC website for details.)  I would simply note here the moment of tautological reasoning wherein ‘the tempo of modern warfare,’ presented as a naturally occurring state of the world, becomes the problem for which faster response is the solution, which in turn justifies the need for automation, which in turn increases the tempo, which in turn, etc.

In elaborating the motivation for the Challenge, Gill Pratt invokes a grab-bag of familiar specters of an increasingly ‘vulnerable society’ (population explosion with disproportionate numbers of frail elderly, climate change, weapons of mass destruction held in the wrong hands) as calling for, if not a technological solution, at least a broad mandate for robotics research and development:

The world’s population is continuing to grow and move to cities situated along flood-prone coasts. The population over age 65 in the United States is forecast to increase from 13 percent to 20 percent by 2030, and the elderly require more help in emergency situations. Climate change and the growing threat of proliferation of weapons of mass destruction to non-state actors add to the concern. Today’s natural, man-made and mixed disasters might be only modest warnings of how vulnerable society is becoming.

One implication of this enumeration is that even disaster can be good for business, and humanoid robotics research, Pratt assures us, is all about saving lives and protecting humans (a term that seems all-encompassing at the same time that it erases from view how differently actual persons are valued in the rhetorics of ‘Homeland Security’).  The figure of the ‘warfighter’ appears only once towards the end of Pratt’s piece, and even there the robot’s role in the military is about preserving, not taking life.  But many of us are not reassured by the prospect of robot rescue, and would instead call on the U.S. Government to take action to mitigate climate change, to dial down its commitment to militarism along with its leading role in the international arms trade, and to invest in programs to provide meaningful jobs at a living wage to humans, not least those engaged in the work of care.  The robot Challenge could truly be an advance if the spectacle of slow robots would raise questions about the future of humanoid robotics as a project for our governments and universities to be invested in, and about the good faith of slippery rhetorics that promise the robot first responder as the remedy for our collective vulnerability.

Will we be rescued?

I’m well overdue for a follow up post to ‘Don’t Kick the Dog’, and the latest media barrage on the untethering of Boston Dynamics’ Wild Cat has got me moving.  An unholy hybrid that runs backwards (at least, in biomimetic terms, according to the bend of its ‘knees’), the Wild Cat appears as a larger offspring of Boston Dynamics’ Cheetah, and a kind of faster, if no less noisy*, cousin of Big Dog (now officially christened with the DOD acronym LS3, for Legged Squadron Support System).

The occasion for Wild Cat’s celebration via a YouTube disseminated demonstration is its release from the hydraulic line that has previously connected it to an offboard power source, into a parking lot where it can dash around seemingly autonomously, albeit under the watchful eye of its minders.  What’s less obvious from this 2 minute demonstration are the enduring problems (or alternatively saving limits) of the power supply, reliant as these devices still are on noisy motors or the short life of batteries.

Also relegated to the margins are the projected applications for the Wild Cat, which brings me to Boston Dynamics’ other recent release, the robot Atlas, and the Defense Advanced Research Projects Agency (DARPA) Challenge that accompanies it.

DARPA-Atlas

As the Boston Globe announces it, university teams selected for the competition are “vying to give [the] brawny robot a brain’ (Bray, August 5, 2013).  Delivered to seven universities selected by the DOD for the semi-finals, the web is now awash with ‘unbox’ videos of technicians sporting crisp Boston Dynamics uniforms unpacking the robot into large warehouses specially configured to receive it, as the select students stand around excitedly anticipating the opportunity to develop its applications. Weighing in at 330 lbs. and standing over 6 ft. tall, the device corporealizes the quintessentially threatening humanoid robot.  As the Boston Globe observes, “It’s easy to imagine Atlas’s hydraulic arms smashing a flimsy wooden door to reach the humans cowering inside.”  But this scene of terror is quickly displaced by the reassurance that “[i]f that ever happens, the humans will probably cheer. Designed by Boston Dynamics, a Waltham company founded by MIT engineers, Atlas is intended to be a first-responder or rescue robot, working in environments too dangerous for humans.”  This fantasy rescue scenario will be carried through in the “real-world” test that will end the current qualifying round, scheduled for December of this year, in which the Atlases programmed by the universities will compete for the chance to prepare for a final competition in December 2014.

Atlas, we’re told, is the DARPA/Boston Dynamics response to the Fukushima earthquake, though it seems highly unlikely that the project was inspired uniquely by, or initiated directly as a consequence of, that event.  The DARPA/Boston Dynamics’ partnership insists that:

“The two-legged Atlas is designed for rescue, not combat. ‘Many of the places where disasters might occur are places that are designed for people,’ said [Boston Dynamics co-founder Marc] Raibert. ‘People can fit in there and maneuver through them.’ So too will Atlas, using the same tools as human first responders. A humanoid robot could climb into a car and drive itself to the disaster scene. Once there, it could open doors, climb ladders, turn valves or throw switches, just like a person.”

Should we be heartened by the fact that DARPA  feels compelled to cloth Atlas’ threatening body in fantasies of rescue?   I’m afraid not.  On the contrary, to my reading this appropriation of the  body of the first responder adds another layer of harm to these projects, resonant with the virtues of the claimed ‘precision’ of the targeted strike.  While DARPA challengers and their sponsoring institutions may find a moral high ground on which to stand under the cover of these reassuring scenarios, the robot’s ancestral inheritance from its military-industrial-academic family is too deeply coded into Atlas’ contemporary lifeworld for the claim of its innocent future to be a credible one.

* Health warning: I would advise using your mute key when watching this video.  It’s also worth noting here the report by Matthew Humphries on Sept 24 of this year that DARPA has awarded an additional $10 million to Boston Dynamics to redesign the LS3 to be “as close to silent as possible so as to be stealthy, while also being bulletproof.”

Robot alerts

One of the aims of this blog is to offer some critical readings of popular media representations of robots, particularly in the areas of warfare and healthcare.  So let’s take the most recent Google ‘alert’ on robots to come across my inbox, dated January 22, 2012.  We get the usual collection of stories, falling roughly into these genre:

Heroic robot ‘rescue’ missions.  Reports on the use of remotely controlled, non-humanoid robots in responding to a variety of emergency situations.  In this case, The Telegraph reports on the use of an ‘underwater robot equipped with a camera’ sent to monitor the area of the wreckage of the cruise ship Costa Concordia in an ongoing search for victims.  A second story in the Irish Independent reports the failure of a Navy team equipped with a ‘robot camera’ to find the bodies of three missing fishermen in a trawler wrecked off the West coast of Ireland.  I note that the almost mundane use of this relatively straightforward technology is performed as newsworthy in these stories through its figuration as at once humanlike, and more-than-human in its capabilities.  A familiar theme, in this case working to keep the robot future alive in the face of a tragic cessation in the recovery of those humans who have died.

Roboticists’ commentaries on the field.  I’m pleased to see Helen Greiner, co-founder of iRobot Corporation and CEO of robotics start-up CyPhy Works, writing a column in the New Scientist urging that roboticists get more serious, less focused on ‘cool’ and more on ‘practicality, ruggedness and cost,’ three qualities that she believes necessary to move robots from promissory prototypes to products on the market. To exemplify the latter she points to the non-humanoid, yet useful Roomba vacuuming robot (perhaps more on Roomba in a later post), and the success of ‘iRobot’s military robots, originally deployed in Afghanistan to defuse improvised explosive devices, [which] proved very useful to the human teams dealing with the nuclear emergency at the Fukushima Daiichi power plant in Japan.’ (See ‘heroic robots’ above.)  Notably absent from mention is the iRobot 710 Warrior.   Nor does iRobot advertise the robot’s ‘firefighting’ potential on its product web pages, but Wikipedia tells us that iRobot has teamed up with Australian partner Metal Storm to mount an electronically controlled firing system on a Warrior, capable of firing up to 16 rounds per second (definitely more on the Warrior in a later post).

Care robots.  The majority of stories echo the pervasive fantasy of the robot caregiver, humanoid projects framed as vague promises of a future in which the burden of our responsibility for those figured as dependents – children on one hand, the elderly on the other – will be cared for by loving machines.  While not my focus here, these stories invariably translate the extraordinarily skillful, open-ended and irreducible complexities of caregiving into a cartoon of itself – another instance of asserting the existence of a world in which the autonomous robot would be possible, rather than imaginatively rethinking the assistive possibilities that a robot not invested in its own humanness might actually embody.

Automata.  Finally, and most interestingly, we find on the IEEE Spectrum Automaton blog a story on the work of animatronic designer Chris Clarke.  Animation, in its many and evolving forms, is an art that relies upon the animator’s close and insightful observations of the creatures that inform his or her machines, combined with ingenious invention and reconfiguration of materials and mechanisms.  Not fetishizing autonomy, the art of animation relies instead on the same suspension of disbelief that enlivens the cinema – some ideas that my colleague Jackie Stacey and I explore at greater length in our paper ‘Animation and Automation: The liveliness and labours of bodies and machines’, soon to be out in the journal Body & Society.