Saturday, April 5, 2014

Back from Doha

We had another successful season testing in Doha, scaling up to bigger teams and testing out new hardware.  

Seven boats in the water at once, in this photo:

George came over and helped out.  He was more help than Alex. 


The staging area.  It was a tough place to have to hang out all day every day.
 


Four boats ready to go.




That isn't an airboat!  The twin propellers help navigate the Arabian Gulf waters much better.



Paul sending a boat on its maiden voyage.  He looks nervous, but the boats mostly performed.  And we didn't lose one.





Nate was working hard on an exciting new user interface concept that allows a single user to control multiple boats at once.  We successfully had Nathan in control of seven boats out in the water.  


Screenshot of the GUI:


Sunday, March 23, 2014

Hakuna Matata

The CRW project recently went to the Mara river basin to help Yale Forestry students investigate the big, lazy bends in the river where hippos hang out during the day.  The hippos, rather impolitely, use the river as a toilet as well as a bedroom.  The hippo "output" builds up in the river, dropping oxygen levels.  Then when floods come through the toilet is "flushed", but the huge amount organic waste can cause oxygen crashes elsewhere in the river and subsequently fish kills.  At least, that is a simple version of the theory the researchers are looking at.

Unfortunately, hippos don't like humans very much, except for brunch, so actually getting into the water to measure it and confirm the theory has so far been infeasible.  That's where robots come into it.  What could epitomize the Robotics Institute's former motto "Dull, Dirty and Dangerous" more than collecting thousands of data points in a massive toilet filled with hippos?  

Our high tech base camp.  We found lion prints about where the photographer was standing.  The quality of food (and the quantity of wine) available at this camp was amazing.


Carefully tying the robot to the top of the land cruiser.  No problem, there will probably be no potholes in the 30 miles of dirt road, right?


Configuring the sensors.   We used a new winch to drop sensors to different depths.  Thanks to RBR and Contour Innovations for sponsoring the sensors.   We were mostly focused on dissolved oxygen, depth and bottom hardness.  We'll post some data when it is processed.


Adding camo.  The theory is that if the boat vaguely looked like a croc, the hippos might just choose to ignore it.  Notice that there are no boats on the Mara, so hippos haven't seen anything like this before.  Given the boat weighs about 18lbs and a hippo might weigh 6000lbs, one only needs to touch it and there will be no more boat.



Putting it in the water.  We took lots of photos of this part, just in case we got some real action.  Notice Abhinav in the background distracting the armed ranger there to protect us.  As many people as possible went to the river edge, using the zebra theory that there is safety in numbers (there were also a lot of conversations about who the fastest runner was, figuring a hippo would only want to eat one of us).  Notice the photographer, yours truly, has taken this photo from a rather higher location.


It sort of looks like a croc, doesn't it?  A croc with a GoPro on its head, to show its friends the dumb humans it had for lunch.


These guys wondered who the relative was, but were too fat and lazy to come and say hello.  Thank goodness. 


A curious hippo watching.  At most of the pools - we deployed about 10 times in the end - they were like this, curious, watching, snorting but mostly just going back under water and waiting it out.  Mothers with babies (surprisingly cute for 1000lb babies!) were the most aggressive, but still stayed away.  


The researchers counted 140 hippos in this pool, but once the boat went out, they mostly went under water and we only saw a head or two now and then.  Would have loved to have the side-scanning sonar!  There were some hints in the basic sonar that we'd gone over a hippo occasionally.


But then there was this pool.  Notice that the boat is on the far right (near the opposite bank), but there are about 10 pairs of eyes closely following the action.  There were more like 20 hippos altogether, the others were under water.   At first they seemed to be pretty chill, just like the others.


But then these two decided they needed to impress a girl hippo.  Or something.  At first they come in a bluffing sort of way.


But then one of them decides that this is serious and that this isn't a croc and it might be tasty.  Or a threat. Or he really wants to impress a girl hippo.  Or something.  We don't know much about what he was thinking.  With their ears and the way they twirl them, it is tempting to think of them as being like Shrek and therefore thinking like Shrek, but there is no evidence to support this.


For 6000lbs in water, these guys are pretty nimble.  They basically run along the bottom.  


Closing in.  We'd been a pretty noisy bunch up until this point of the trip.  But apart from camera clicks and Chris Dutton throwing a stone (the rangers would do this, too, apparently it is often enough to dissuade a croc or hippo, unfortunately Chris is not Drew Brees and I'm not sure his stone actually reached the water.  Next time we'll bring Drew Brees, or maybe Brett Favre, he seems tough), there was not a sound, no one even breathed.  It was an amazing, scary, incredible sight.  None of us will forget it for a very long time.


 But some good driving by John (by good, I mean go as fast as possible in a straight line and hope), meant we got away.   Amanda and Chris, the water researchers, think that hippo couldn't run in the deeper water.  The didn't think we'd get away if the hippo had been able to keep running.   For the rest of the trip, we made sure we knew where the deep water was - partly thanks to our Lowrance fish finder and ContourInnovations maps (how's that for a sponsor plug?!).



Actually, John deserves some credit for good driving.  Another 30 yards and we were off down some rapids.   Sort of sounds like the Indiana Jones version of field robotics.


After some discussion we realized we had no choice but to drive back past the hippos to be able to retrieve the boat.  It only occurred to me when I looked at this photo that we were encouraging the hippo to run at us?!  Fortunately, apparently hippos have the stamina that it looks like they have and they just watched it go by.



The East Africa office of the World Wildlife Fund came down to check out what we were doing and got hands on, helping clear a clogged cooling tube. 



Actually, there was a lot of interest from the locals.  The monkey on the branch followed the boat a good 50 yards, hopping from tree to tree then found a scenic spot and watched carefully.  We suspect he was looking for revenge for kicking him and his buddies out of the land cruiser.


Not everything went smoothly and not everything was easy, but the guys did a great job, collecting a lot of data in several locations, sometimes going back multiple days to get data before and after heavy rains.   We hope to get back to Africa some time and work some more.


Thanks to all the sponsors/helpers: Project Olympus, Contour Innovations, QNRF, RBR, CMU-RI and Yale Forestry.  Very special thanks to our amazing hosts Amanda Subalusky and Chris Dutton from Yale Forestry.

View most of the above in higher quality in this pdf file.


Wednesday, November 20, 2013

System Characterization

How some of us spend our nights at Platypus - working on characterizing our system a bit better so we can improve future hull iterations, design more power efficient algorithms, and update our controller. The end result will hopefully mean longer operating times, improved control, and less operator time saving boats from entanglement in shoreline vegetation.







Sunday, November 17, 2013

Depth Mapping

Platypus has started working with Aquatic Edge Consulting and Contour Innovations to produce high-quality, low-cost depth contours and vegetation maps, like the ones below.







Wednesday, October 30, 2013

Platypus on The Verge

The Verge website came out and filmed Platypus and the airboats as a part of short film about Pittsburgh's transformation from steel city to robot city.

Check it out.

Monday, October 28, 2013

Platypus goes to China!

Paul recently traveled to Nanjing to begin working with NUST and the airboats.  Zesheng and Yaonon did most of the work making the boats work in a beautiful NUST pond.





We also showed the boats at a robotics exhibition at a huge new technology park just outside Nanjing.

We are looking at the possibility of opening Platypus-China with researchers from Nanjing University.  Specifically, Platypus-China would look at shrimp and sea cucumber farms.