1)Why is the Navy firefighting robot Saffir bipedal?; 2)Word for robot/human interaction?

Factual Questions
1

See subject.

Bipedal is hard.

The Navy, in this public rollout of a prototype, say that it is designed specifically for the service environment–that is, service in the physical/hard environment–and with their human/soft environment–which naturally includes the humans working with it (ergonomics) and, in a twist, at least publicly stated, their morale.

Seems to me an extraordinary effort that both could be addressed with different designs, as to the physical environment–certainly mobility. As for the anthropomorphism, it seems unnecessary; the morale/interaction of soldiers with their machines/tools has always been quite robust and adaptable.

Also, there must be a word or various words like “ergonomics” that roboticists/psychologists–in reality or certainly in sci-fi–have for the specialists who design for social/thinking interaction. Any help?

2

At first glance, it seems a very rational choice: a bipedal robot with human dimensions can go anywhere a human can go, climb the same stairs, etc. So you don’t have to redesign every ship and submarine to account for a firefighter that looks like Wall-E.

3

The environment s heavily optimized and was designed with soft squishy bipeds in mind. “Doors” are designed for stepping up and over a pronounced lip. Stairs are the norm for moving between levels and they are steep and cramped compared to normal civilian building codes. That environment is generally a carefully crafted billion plus dollar moving facility that has to deal with the real challenges of someone intentionally trying to blow holes in it while with large quantities of explosive and flammable materials.

It may well be more efficient to craft a robot that fits the existing environments than rework the environments for a cheaper/simpler/less challenging robot design.

4

In answer to the second question: “Social robotics”?

5

I read the story on CNN and the Navy addressed this. As others have noted, ships are built with certain features that work great for people with legs but not so well for wheeled or tracked vehicles (and robots, I guess) - hatches with tall lips at the bottom, steep ladders, etc. They also said that building the robot in humanoid form allows it to use firefighting gear that was built for people (both things like hose nozzles as well as some protective gear).

6

Look, the M-1 Abrams is a great tank. But it’s too heavy to cross some bridges and it’s too wide to operate in some urban environments. Same sort of thing for Naval firefighting robots. And they also need a place to store them in readiness yet out of the way when there isn’t a fire that needs to be put out.

7

Look at a similar real world example. Have you ever stayed in a hotel room designed for handicapped (wheelchair using) folks? I have, and it’s a PITA. Nothing fits well; *everything *is in the wrong place. Which I recognize also means that rooms designed for folks with two good legs are at least equally a PITA for wheelchair users, if not much worse.

The difference there is comparatively minor. Feet-using adults are 5-6 feet tall, while wheel-using adults are 3-4’ tall. Wheel-users are also wider and longer, more like 3’x4’ vice 2 x 1.5’. And wheel users are much less tolerant of lips and ridges on the floor.

Other than those small points wheel-users and feet-users have identical sensors, communicators and effectors. IOW eyes, ears, talking, listening, and arms & hands work exactly the same in both models.

The point being that even for this fairly minor difference in shape, the built environment that fits one type well fits the other type poorly.

With that in mind …
Naval ships are much more highly built environments that place much *higher *demands on user compatibility. It makes *complete *sense to build machines that are the same shape as humans and can hold and use designed-for-human tools. It’d be stupid to do otherwise if the technology permits anthropomorphism.

That can be carried to far, putting cute “faces” on these things. But at the same time, cameras mounted at human-eye level make sense so the robot can easily see indicators designed for humans. But that’s not a reason to not also install cameras on long flexible stalks like a snail, or on the feet, or whatever additional non-humanoid extra capability would be useful for the task. So the robots don’t necessarily end up resembling humans in all respects, just most of them.

8
  1. Cybernetics?
9

The stairs are called ladders as that is what they resemble.

That SAFIR was tested in an open and flat space. They need to get into one of the main spaces or a passageway with numerous knee-knockers to step and drag hoses over. Some places, are barely high enough for someone 6’ to walk in. They should make the robot 5’, no more.

I also wonder how it works in a non-level environment. The ship will begin to turn in a circle and therefore slightly roll to one side. You do this to control the excess water that is created. You might want to drain the water away or have it stay in the area.

10

Yes, that’s key.

11

It still seems to me that four feet might be better than two. You can still use them in pairs, or only use two of them, for things like ladders, but on flat ground, balance would be a lot easier. Or heck, you could design the extra legs so they could be swung upwards, to use like we use hands on a ladder, leaving the purpose-built hands free to continue to carry hoses or other tools.

12

Like this? http://www.digitaldoodles.com/files/mini_gallery/ROKSTour_DestroyerStairs_1465.jpg

Are people expected to carry heavy objects up and down these or is that handled through some other means?

13

The current DARPA Robotics Challengealso involves designing/building a robot to perform disaster recovery in an environment built for human beings, and most of the entrants were humanoid. There were/are some more innovative designs, but generally when opening doors/climbing stairs/crossing rubble-strewn ground there is some definite advantages to the human figure.

The DARPA Robotics Challenge is also designed to push the envelope of robotics design, of course, so they are not looking for an off-the-shelf solution to a specific problem.

Also, when I worked in the field we typically called it Human/Robot Interaction, but we were better at engineering than coming up with names for stuff.

14

Very typical, I’ve lugged all sorts of gear up and down ladders like that.

I was a Damage Controlman Second Class. Fancy way of saying a shipboard firefighter. I’ve been up and down ladders like that in full turnout gear and O2 rig and a heavy ass barely flexible charged hose.

Short of exploding and sinking in minutes fire at sea is the scariest thing there is. If your house is on fire you can run out. In a house fire the firefighters can evacuate at any time if its too bad. Shipboard firefighting is very aggressive. Save the ship at all costs. In case of a major engine room conflagration the first couple of guys in the door were considered expendable. (Ah to be 20 and fearless again.)

I wish them all the luck in the world with the robot.

15

So much for tracks and wheels. But why not four legs, rather than just two? Four legs can do anything two legs can.

16

Didn’t George Orwell say that?

[I was dying to work that in OP, but edited it out.]

17

Bipedality is not an advantage, but a necessary sacrifice. The only reason why we walk on two legs is that we’re limited to a total of 4 extremities. Nature can’t just replace an extra pair of legs that have been turned into hands - at least not without turning those hands back into legs. But robots don’t have that limitation. Why should they suffer the disadvantages of bipedality?

Why 4 is better than 2:

  • The legs can be less sturdy, given that each of them has to only carry 1/3 of the body at normal gait, and 1/2 when running. Conversely, legs belonging to a set of 2 have to alternately bear the full weight of the body they’re attached to.
  • 4 legs are inherently more stable.
  • Stability can be achieved with minimal brain processing power, whereas bipedality requires many calculations per second, just to keep the body balanced on its two feet, even when just idly standing around.
  • Stability can be maintained without any expenditure of work, allowing the entity to rest, hibernate or even completely shut down. Bipedal entities must constantly shift their body weight around to maintain stability, which consumes energy - or must have a very low centre of gravity, making them short and bulky - or have huge feet, making them unable to negotiate stairs and narrow turns.
  • 4 legs can be spread apart and increase stability in all directions, allowing the body to pivot and extend anywhere, with minimal risk of unbalancing. 2 legs can only spread along a line; stability is increased in one dimension, but remains small in the other.
  • Losing 1 out of 4 legs due to damage or breakdown does not necessarily cripple the robot to the point of immobility; losing 1 out of 2 legs, however, leaves the robot stranded.
18

Did I kill the thread? :dubious:

19

The 2015 DARPA Robotics Callenge ended yesterday. The best in humanoid endeavor. Wheaties boxes and Kardashian and magazine covers are in the future.

Taking first place and the $2 million in prize money that goes with it is Team Kaist of Daejeon, Republic of Korea, and its robot DRC-Hubo. Coming in second and taking home $1 million is Team IHMC Robotics of Pensacola, Fla., and its robot Running Man. The third place finisher, earning the $500,000 prize, is Tartan Rescue of Pittsburgh, and its robot CHIMP.

[snip]

Launched in response to a humanitarian need that became glaringly clear during the nuclear disaster at Fukushima, Japan, in 2011, the DARPA Robotics Challenge consisted of three increasingly demanding competitions over two years. The goal was to accelerate progress in robotics and hasten the day when robots have sufficient dexterity and robustness to enter areas too dangerous for humans and mitigate the impacts of natural or man-made disasters.

The DRC Finals competition challenged participating robotics teams and their robots to complete a difficult course of eight tasks relevant to disaster response, among them driving alone, walking through rubble, tripping circuit breakers, turning valves and climbing stairs. A dozen teams from the United States and another eleven from Japan, Germany, Italy, Republic of Korea and Hong Kong competed in the outdoor competition.

The DARPA site for the Challenge. Just look at the gallery of entrants for starters…even designing the robotics course, which took years, and getting the national standards people interested, is fascinating.

A speeded-up film of the winner doing its stuff is here. It’s extraordinary. I hope OP is suitably ashamed of himself.

DARPA has its own Youtube channel, with real-time 12-hour recordings of the live broadcasts of the robots slowly doing their stuff, falling, and the rest. Hell, a panel on the Challenge is being broadcast now. [Note to future zombie readers:* nyah, nyah*…]