Anna Konda, the Robotic Firefighter

Roland Piquepaille writes "In fact, Anna Konda is a robotic fire hose moving like a snake. This robot, which has been developed in Norway by SINTEF, is 3 m long and weighs 70 kg. The snake contains 20 water hydraulic motors that move the robotic joints. And the energy needed to power these motors comes from water pressurized to 100 bars and already available inside the fire hose. This gives enough energy to this water-powered robot to climb up stairs, to lift a car up off the ground or even break through a wall. Very clever design! The designers think that this robot could not only replace humans to fight fires when it's too dangerous for them, but could also be used for subsea operations or explosion prevention. An additional overview contains more details and pictures of this snake robot."

Comments

Roland the Plogger again - here's the real link

[Animats]

Here's the real link to the research. As usual, Roland the Plogger is posting a story from a blog, maximizing ad revenue, and the actual reference has been lost. One would think that Slashdot's "editors" would be wise to this by now, but they still don't get it.

It;s only a prototype; the water stream that comes out is more like a garden hose than a fire hose.

And Then Again, Maybe Not

[MikeyTheK]

It's in interesting idea, but it also has the usual drawbacks, namely it is extremely difficult to move 200 feet of 2-1/2" for a team of firefighters when the line is charged - the weight is not the only problem. A charged line is stiff (ha, ha), so moving it arount corners, into rooms, up and down stairs, etc. is very awkward. You can't just drain a line every time you want to move it. It takes too long. So, you normally have to move a line while it's completely empty (called "flaking" the line), then charge it, or after it's charged, fight with the line the whole way. On top of all that, it's very easy to kink or even twist and decouple hose, which is, of course, disasterous. Normally what we do is carry as much line as we think we are going to need to a safe zone of the structure we're fighting, then flake it into big loops right there. That way we have all the hose we need in one place, and we can just extend into the hot zone without kinking, and also dragging the charged line the minimum distance to do the work we are going to do.

There is also the other problem: We typically charge our lines to the point where the nozzle-man's feet just leave the ground, then we ease back so they are just barely on the ground. This maximizes our flow into the area we are fighting. With a two-person nozzle team that means we have in the neighborhood of 600-700 lbs of ballast to offset the reaction force at the tip (the force of the water exiting the nozzle that is pushing back on it, which would result in the hose flying all over the place otherwise). (The 600-700 lbs is the weight of two firefighters, their bunker gear, air packs, etc.). The robot only weights 70kg, so it won't have nearly the control of the tip, which means that you can't push nearly the water.

I could see this as a good application when trying to work in a warehouse or supermarket, where the distance from the door is large. However, the device is going to need assistance to move a great distance since the line has to be charged in order for it to function, but if the line is charged it becomes much harder to move the line. That combination seems to defeat the purpose - of keeping firefighters out of harm's way.

Personally I'm in favor of our current option "b" - trench cut the roof (long cut perpendicular to the path of the fire, in an area we know is good), then drown the cut with water from a ladder pipe, which causes a water curtain - the good part of the building is saved by the water curtain, which means we can fight the remainder from a position of strength.

Re:And Then Again, Maybe Not

[idontrtfm]

Underwater or space is the most likely place for robotic snakes. I did enough work with them in grad school that I believe weight is going to be a big problem at least during my life time, and certainly during my grad life it was. Researchers very often have pie in the sky stories like this to tell, but I do not think them practical.

Re:And Then Again, Maybe Not

[SmurfButcher Bob]

Hey, if they can make a deuce-and-a-half that can get itself up 3 or 4 flights of stairs... I'm all for it. If they can make this thing go *down* a flight of stairs into a basement... I'm all for it.

There's just a couple of caveats, however -

1. Typical hose lengths on an engine (in the US, at least) are 150', 200', and 400', made up of either 50' or 100' segments. There's no way in hell the entire length of hose will contain a "snake" exoskel as someone had suggested... meaning you'd probably have several snakes separated by "normal" hose lengths. This stuff needs to fit on the truck, somewhere... and a mere human needs to deploy it WITHOUT a Genie-lift.

2. The unit only operates when the line is charged. As Mikey said, that means the entire length of the line is rigid; typical fire streams run at +60 psi (plus additional for friction loss and altitude) for a smoothbore, or up to 160 for a "taskforce" (fog) style tip. Such pressures are not condusive to bending. Combined with having a coupling every 50' along the length of the hose, which invariably get snagged on every corner or edge they find... the idea of this device pulling the hose behind it has some serious challenges.

3. Speaking of bending, it'd be interesting to see how this device would be stored on the truck. A flat-load is probably out of the question... and space on a truck is typically at a premium. I assume the "snake" would retain it's circular cross-section, compared with a typical hose which squashes flat.

4. Somewhere, there was mention of a camera at the head of the unit. Since your typical CCD is useless in a smoke condition... your starting price for vision is $13k for a bolometer, and that's just the sensor. By the time you've added everything else needed for vision, including the "Motorola Floating Decimal Point", remote-vision will probably top $25k, easily. (And yes, I'm pulling numbers out of my butt. However, your typical bolometer-based TIC will run about $16k-$23k for a hand-held unit, and those are mass-produced. Remove the production volume, separate the sensor from the display by 200', realize that "wireless" is not an option, invent a cabling system that'll withstand 1000+ deg...) I don't see anyone who could afford to run this thing via remote vision.

5. I *WILL* disagree with Mikey on the Noz-Reaction argument. He's right, but sadly he lives in TFT-land:

5a. Water weighs 8lbs per gallon; whatever your reaction force is, you need merely have adequate water-weight in the hose on the ground, behind you. (Yes, that description was awful). I weigh 130lbs soaking wet, and can solo a deuce and a half at 170psi with a TFT. Don't ask me to move it while it's flowing water, but once in position... I can solo it without any effort, and maintain control over a 30 degree arc. The trick is to have the line continue *straight* behind you for a few meters, such that the hose jacket takes the compression from the reaction force; meanwhile, residuals keep the jacket from kinking, and mass & ground friction prevent the hose from sliding backwards. It's not appropriate in many places (e.g. if a tight corner is involved), but it's effective in others. Barn fires and corridors, for example :)

5b. Pressure isn't volume, especially with a Thinker nob. Most thinkers intend 120lbs (or whatever) at the tip, and they clamp anything above that. Remind your guys that the typical Taskforce Tip has Thinker built into it, so the excess pressure is NOT increasing the GPM; you crank it up, the Thinker chokes it down at the bale. Most TFTs don't have a visible control for this device, but many Master devices do. Take a look at the deck gun on your engine - it probably has an adjustable GPM collar. So long as you meet the minimum pressure for the tip, it doesn't matter what pressure you dump into it - if you set the collar at 900 gpm, you'll GET 900 gpm. It won't matter if you're at 130psi or 250psi... you'll get 900gpm. Most TFTs contain this device (it's a selling point), but it's f

Re:And Then Again, Maybe Not

[gtwilliams]

to get from 99 to 100 takes... wait for it... 1400 calories. That is a LOT of heat.

Actually, that's too much. The heat of vaporization of water is 539 cal/g at 100 degrees C.
But, of course, the concept is dead on -- to remove a LOT of heat, turn water to steam.

Re:And Then Again, Maybe Not

[SmurfButcher Bob]

Great insight - but there's nothing to test... there are several other ways of bettering the conversion rate without the base temperature being 99deg. And, we wouldn't need to deal with a hose-line that is just below boiling.

1. Surface area via HPF. High Pressure Fog systems were all the rage a few decades ago, and also in the 1800s (wait long enough, everything comes around again, eh?). The conversion rates are dangerous if the application rate is not properly managed. Likewise, application is dangerous because fog doesn't travel; you physically need to be next to the hot spot. Not only does this expose the human to untenable radiative heat, but the resulting steam production will instantaneously boil any persperation in the shell of his gear, blowing it directly onto his skin underneath. HPF quickly lost appeal in manned application scenarios because humans are not compatible with it; any gear capable of protecting the user from the radiation and resulting steam bath was not and still is not capable of surviving the intense crawl from the front door to the fire room. It certainly will not survive it twice, and neither would the person wearing it after they collapse from heat exhaustion. HPF still has strong promise of making a return, however, because it is perfect for mitigating atmospheric threats (rollover / smoke explosions) without inverting the strata.

2. Surface area via aeration (Compressed Air Foam). CAF is a low pressure method of placing "bubbles" into the supression agent (water), consisting of a typical Class A foam agent and compressed air injected after the pump. The result is a very fine, uniform foam - like soggy shaving cream, pure surface area and little volume. It also sticks to things, as opposed to ending up on the floor. The surface area allows for an explosive conversion rate, and is dangerous (severe steam burns to the crew) if proper expansion paths are not created to route the steam flow away from the users during application. Application technique is further complicated by needing to close all vent paths after the application is complete. The Class A agent likewise acts as a sofactant, lowering surface tension and allowing penetration which is a huge advantage over HPF... HPF is useless on a burning couch, mattress, or hay bale for example. The stickiness allows it to be used as a heat-conduction insulator for protecting unburned regions, and an IR absorber for slowing flashover. If the compartment can be adequately sealed up, direct application to the heat source is not necessary; the agent can be applied anywhere that the smoke path / IR path will produce an adequate conversion. Unlike HPF or straight water with a TFT, CAF application is not effective at preventing smoke explosions / ignitions (rollover) in the thermal strata... so if one starts while you're under it, you're screwed. CAF is most effective when applied from outside the compartment just before flashover, or can be applied inside the compartment in a post-flashover environment. In short, the more heat and pressure inside the compartment... the better your performance will be. CAF is significantly less effective in open / well ventilated areas, but still beats the snot out of foam or water - with potential conversion rates up to 90+%, as opposed to 3 or 4% with water or foam. Typical kitchen fire with 1200F degrees at the ceiling: 800 gallons of water directly applied to the fire to suppress, 20 minutes for the ambient temperature to drop to 200 degrees. Same fire with CAF: 80 gallons applied to the ceiling (ABOVE the fire, not ON the fire) by users outside the structure. Temp drops to 250 degrees in 45 seconds. After vents are created for the steam to exit, temp drops to 200 in another 7 minutes.

So, while near-boiling water may add a few points of efficiency to the conversion, it's kinda moot... we've got the efficiency for those cases we can use it. If victims are involved, for example, we cannot use it. We also cannot use it if we'll be sitting the the resulting steam path.

Re:And Then Again, Maybe Yes

[arthurpaliden]

Keep the hose partially charged when moving, to stop kinking, then max the pressure when you get it in position. This would be a great asset when fighting shipboard fires where you do not have the optional advantage of fighting the fire from outside the structure.

Translation of Norwegian Dagbladet.no article

[ThJ]

Here's my translation of the Norwegian Dagbladet.no article.

The Worlds Most Sophisticated Firehose

(Dagbladet.no): Tenk deg en situasjon hvor du er innestengt som følge av brann, snøras eller jordskjelv, og det er for farlig for hjelpemannskapet å gå inn og redde deg. Det er da Snakefighteren Anna Konda kommer glidende inn, på skrå sidelengs som en ørkenslange.

(Dagbladet.no): Imagine a situation where you're trapped due to a fire, snow avalaunch or earthquake, and it's too dangerous for the rescue crew to enter to save you. That's when the Snakefighter Anna Konda comes gliding in, sideways like a desert snake.

Hun er verdens sterkeste og mest avanserte brannslange, ifølge SINTEF-forskerne Pål Liljebäck og Øyvind Stavdal som har utviklet henne.

She's the world's strongest and most advanced firehose, according to SINTEF researchers Pål Liljebäck and Øyvind Stavdal, who developed her.

En Anakonda av metall, smidig, sterk og smart, inspirert av naturen selv. Hun kommer hun seg frem gjennom alt slags terreng, og har sanser som en vanlig slange ikke har.

An Anaconda of metal, agile, strong and smart, inspired by nature itself. She moves through all kinds of terrain, and has senses that a regular snake lacks.

Hun kan heve hodet og sprute vann, slå gjennom vegger med en slagkraft på 700 kg i tyngdefeltet, løfte vekk objekter for å frigjøre fastklemte dyr eller mennesker og bringe gassmasker. Ved hjelp av infrarødt kamera, ultralyd og sensorer skal hun kunne finne kropper og kartlegge et område.

She can lift her head and spray water, break down walls with a gravity of 700 kg, lift away objects to free trapped animals or people, and bring gas masks. With an infrared camera, ultrasound and sensors, she'll be able to find bodies and map an area.

- Snakefighteren representerer en ny æra i brannslukning, sier Pål Liljebäck ved SINTEFs avdeling for anvendt kybernetikk til Dagbladet.no.

- The Snakefighter represents a new era in fire extinguishing, says Pål Liljebäck at SINTEF's department for applied cybernetics to Dagbladet.no.

Han presiserer at slangen er et verktøy, ikke en erstatning for brannfolk, for hun er ikke skapt for å dra med seg objekter.

He notes that the snake is a tool, not a replacement for fire crews, since she wasn't created for towing objects.

Vannhydraulikk

Anna Konda drives av vannhydraulikk, som er nærmest et ikke-eksisterende fagfelt i dag, ifølge de to forskerne. Det vil si at det er vannkraft som driver musklene til den tre meter lange, 16 cm i diameter tjukke og 70 kilo tunge slangen. På et brannåsted kan hun tilkobles en brannslange. Er det snakk om en sammenrast bygning, kan hun ha en innebygget dieselmotor og ha med eget vann. Hun beveger seg med en hastighet på 20 - 30 cm i sekundet, men målet er en fart på en meter i sekundet.

Anna Konda is powered by water hydraulics, a virtually non-existing field today, according to the two scientists. This means that water power is powering the "muscles" of the 3 meter long, 16 cm in diameter thick and 70 kg heavy snake. In a fire location she can be connected to a firehose. In the case of a collapsed building, she can carry her own diesel engine and her own water. She moves with a speed of 20 - 30 cm per second, but the goal is a speed of 1 meter per second.

Slangen skal dels fjernstyres, dels ta egne avgjørelser. Hun bruker så kompliserte bevegelser at hun selv må greie føle seg frem i terrenget og beregne hvordan hun skal ta seg frem. Men en operatør skal kunne gi overordnede instrukser.

The snake will partly be remote controlled, partly make her own decisions. She uses such complex movements that she has to feel herself through the terrain and calc

Want pictures?

[scdeimos]

Due to the complete lack of pictures in the source article, here's some for your appetite...

Anna Konda in action (JPEG, 844x453)

Close-up of a segment (JPEG, 280x210)

Other snake 'bots

[HoneyBeeSpace]

For some other snake robots, check out these links:

http://www.snakerobots.com/

http://arctangent.8k.com/snake/snakemain.htm

Re:100 bars?!

[SmurfButcher Bob]

Go back a bunch of decades, to the "High Pressure Fog" days. Lots of engines produced that pressure, and they used 1" boosters.

Also, remember that this is a Euro invention. Think it's gonna have a smoothbore? A TFT?

Nope... it'll be HPF with a 1". Europeans have been begging for a way to bring back HPF for almost a decade... and this might just be a viable way to do it. Expensive as hell, but viable.