Keeping animals from tropical regions of the world in a cold climate is an expensive business, they need a warm environment in their pens and sleeping areas. Marwell Zoo was spending a small fortune keeping its herd of nyalas (an antelope, not as the title suggests a deer, native to Southern Africa) warm with electric heating, so they went looking for a technology that could reduce their costs by only heating while an animal was in its pen.
One might expect that a passive IR sensor would solve the problem, but a sleeping nyala too soon becomes part of the background heat for these devices, and as a result, the heaters would not operate for long enough to keep the animals warm. The solution came from an unlikely source, a coffee queue monitoring project at the IBM Watson headquarters in Munich, that used an array of infra-red sensors to monitor the changing heat patterns and thus gauge the likelihood of a lengthy wait for a beverage.
In the zoo application, an array of thermal sensors hooked up to ESP8266 boards talk back to a Raspberry Pi that aggregates the readings and sends them to the IBM Watson cloud where they are analyzed by a neural net. The decision is then made whether or not a nyala is in the field of view, and the animal is toasted accordingly.
This project has some similarities with a Hackaday Prize entry, automated wildlife recognition, in its use of Watson.
Nyala image: Charlesjsharp [CC BY-SA 4.0 ].
Yum, toasted antelope.
I was thinking roasted would be my preference :)
My first thought also, mmm-mm-mmm.
mm IBM Watson, the big corporate machine learning scam
do you have a story to tell?
simply google “IBM Watson scam” and read first page results
The results say “rasz_pl is too lazy to explain himself”. Sounds about right.
And what happens if you Watson “Alphabet Google scam”? [1]
;-)
(full disclosure: I consider both of them somewhat of a scam, just on a different level)
For those who don’t get it: note that Google has skin in the game, and asking it whether a competitor is good or bad might deliver skewed results. And vice-versa. Think for yourself, etc.
I’m Duckduckgoing it. Trying to get into the habit of using it. They seem like decent respectable chaps, at least from their own spiel.
Or, you could save some money by getting a few of those super powerful mil surplus lasers tied into a sentry gun type carriage.
One shot and your cold nyalas are a thing of the past, isn’t the future supposed to be benefiting from uploading everything to ‘the cloud’?
I wonder could a few more sensors such as few IR lasers and ultrasonic sensors or even just a pressure sensor like what you see on some automatic doors allow them to get rid of needing Waston and make do with just a Raspberry Pi or even an 8bit uC?
For fast response maybe use an IR quartz heater or just an array of incandescent heat lamps.
but that wouldnt land cross promotion deal with IBM. Those $x-xx mil a year IBM support contracts dont sell themselves. Just be glad they didnt have to roll out the big guns, like blockchain accelerated animal detection.
They could easily heat the nyalas with the power ~wasted~ used to do the blockchain computation. It’s win/win!
I was thinking maybe a camera, the Pi can do image recognition right? Nyalas are apparently brown-orange, so have the pen some other colour. Big enough blob, Bob’s yer uncle. Nobody says you have to use infra-red.
But now I’ve read Rasz’s comment. Yeah. Uploading to “the cloud” (which always seems to mean “one company’s server), and getting a supercomputer to thermally image an animal from random IR sensors… apart from making use of an old IBM novelty nobody wants any more, what’s the point?
They’d be better using a pressure switch under the Nyala, and using Watson to generate the heat.
This is the opposite of a hack. It’s not clever, neat, or taking advantage of things from a different angle. It’s throwing ludicrous amounts of hardware at a problem that basically needs a switch. For media publicity. Shit story, HAD. Of course I’m sufficiently reassured to know what nobody paid anybody for this story to occur.
“They’d be better using a pressure switch under the Nyala”
^^ this
Geez, talk about overcomplicating a simple problem. Just lay down pressure sensitive burglar alarm mats !
To be honest, they probably do. Unplug all the Watson and Wifi stuff once the photographers go home (do journalists actually go look at things themselves these days, or has the miracle of press releases freed them from that tiresome burden?). Then put back the burglar alarm pad, relay, and heatlamp, that one of the zookeepers had put together in his shed.
How much energy does Watson use compared to leaving the heater running?
There are surely simpler ways to solve this problem
It has something like 2880 cores and 16 terabytes of ram so probably a few tens of KW though the power 7 architecture uses much less power than an equivalent X86 system.
Switching to natural gas for the heat source would reduce your always on cost by at least 90%. If you want the system on only when the animals are present, I’d suggest pressure or capacitance pads on the floor in combination with a PIR sensor. That should be good enough and much less likely to fail than this rube goldberg machine that have going now.
I really don’t under stand there problem.
There are so so many ways to keep them warm with out using so much computing power and tek.
Well ti was a read for me on a sunday morning.
2 sensors at the door? Animal goes out heat goes off,animal comes in heat goes on. Parts salvaged from a couple junk garage door openers….