EMF in pipes and Mixing in Space

I was thinking that it would be cool experiment to try to see how much current would be produced by flushing magnetic particles through a system of pipes that were wrapped with wire and connected to batteries.  Wonder how powerful the magnets would need to be and how the fact that they are traveling in water effects the field.  Would be interesting to compare to tap water to see how much an emf field is present already.

Just not sure if it would work because:

  1. the magnets would all clump together maybe not due to attraction and repulsion maybe end of
  2. field not strong enough to induce a noticeable current

 

[update] Well I I talked with someone that knows a bit more and he reminded me that even though it might be possible the magnet would be slowing down the water and therefore not doing any good.  He suggested that a mechanical method would be better rather than using magnets as well.

Was wondering how do you mix liquid things in zero gravity?  I would imagine it is either really hard or not possible.  Maybe we could use something like http://phys.org/news175873414.html.  I guess we could just shake them together…

Well as to be expected they have done some experiments and they tried mixing oil in water by shaking them together.  The cool thing is that they stayed together for 10 hours!  On earth they would separate in 10 seconds.  They also tried mixing water and water that had tea in it by pouring them into the same container.  However, since no gravity they barely mixed at all.  However, they did not shake the two.  However, I’m assuming that they would mix and would not separate.  (source: http://history.nasa.gov/SP-401/ch12.htm)

Trash Water Sewer oh my

Well, I was thinking that  a fun multirobotics project would be to have robots that were trash cans and their goals would be:

  1. Minimize expected time to find a trashcan
  2. Stay out of the way
  3. When full empty into “dumpster” or something
  4. Stay charged
  5. Minimal movement (don’t move when someone is using it)
  6. (Some sort of tipping defense mechanism?)

They might be outfitted with the ability to identify and suck up trash (like a vacuum).  They would have to coordinate so that they did not cluster.  They would learn roles and adapt to changes in the number of robots.

This multirobot trash can could be used at high traffic events or locations.  I guess the big thing is that people might actually steal the trash can!  We would need body guards for the trash can haha ;).  I do realize that this is sort of like the room clearing problem for multiple robots, so I’m thinking that some techniques that have been developed could be applied to solve this problem.

I call them CoWs, Cooperative Waste-cans :).

So, water and sewer plants when there are multiples that must cooperate to serve a community. Then the plants would be the agents and they could each communicate with the others to coordinate how/who/when they would do things.  This would reduce risk by distributing the system.  It is also agent based because the plants are selfish.  It is not a distributed problem because the facilities may be different companies.

They mush cooperate to take better care of the environment (over taxing the river where the water comes) and customers.  I wonder who owns the sewer pipes then?  Would this be a traveling repairman problem?  So, it would have to be distributed due to the joint ownership of the pipes by different companies, but separate management of the repairmen?

In the creation of the city that would also be an interesting problem of determining all the sewer layout and design…  This seems like a Ford-Fulkerson solution with a Traveling Repairman problem and TSP.  So, essentially we could evolve different pipe networks so that when the roads get built on top we will have created the best possible network with the given simulated info at the time.

 

Repairmen and Trees

I think I have finally found a problem that has not been studied and is important.  I’m calling it the K-D-Traveling Repairman Problem.  K, the number of repairmen, D, the number of types of repairmen.  The repairman problem is a NP-hard problem and therefore heurstics are used to solve it.  The Predictics group at CMU used Machine learning to approximate solutions to the TRP.  I’ve found one other author that has looked at the K-TRP problem.  I can’t find anyone that has looked at the K-D-TRP.  This is a heterogeneous extension.  Basically I’m saying that there are different types of repairmen that may need to overlap at nodes such in order to perform the repair.

Another interesting extension is a dynamic graph.  Also, I’m interested in looking at saturation levels of repairmen.  Non-deterministic repair times would also be another aspect to look at.  The nice thing with this problem is that it is based on a real world problem that really needs to be solved.

Also, want to look Game theory, optimization theory and expander graphs seem like they will be useful in the theoretic aspect of the problem.  I want to then develop a multi-agent learning algorithm where the agents learn how to solve the problem.

Also, I have been reading about some cool trees: KD-Trees and KD-B-Trees.  They are pretty cool.

Bribes

Another idea is to use bribes to manipulate the system http://mpref2012.lip6.fr/proceedings/MaranMPREF2012.pdf show how to resist bribery.  Can we use what they learned and instead use bribes for good as a means of inducing cooperation?  The problem trying to solve is how to bring about cooperation of self-interested agents when the only/main reason they would help is because you are bribing them to do it.  This is a weak form of cooperation, but could still produce cooperation.  Bribery is usually used to have someone do something illegal, so one example would be to bribe members of a team to give you information that your team could use against them.

I noticed the other day that this is a topic in the MAS book page 323.

Kramer Middle School Botball

I’ll write about this then…

Well I published that above on Feb 21 2014 and now its May 19 2014!  I am just now taking the time to right about how things went.  It was a busy semester.  So, I’ll just go right into it.

So, as part of CS 880 (Multi-Robotics) we volunteered at one of three underprivileged middle schools (initially 4 but one of them pulled out) Jefferson Houston, Kramer and TranSTEM Academy (which is I think is Cardoza) to teach robotics and prepare a team to enter the Botball competition.  I got to volunteer at Kramer Middle school, since while at the initial Botball workshop I worked with that middle school teacher and student.  At Kramer Middle school I worked with Jesse Hatfield (another PhD student that was in CS880).

Things that were interesting about Kramer:

1. “99% Black and 1% mixed race” (http://profiles.dcps.dc.gov/Kramer+Middle+School).  The first time I went to the school one of the students was like “Look there’s a white guy!”.

2. The students wear a uniform White shirt and black/dark blue khakis.  I didn’t notice this until about the third time I went.  Interesting, I never had to wear a uniform and I went to public school in middle school…

3. They separated the classes by gender.

So, back to botball…  Our primary contact teacher at Kramer was Ms. Jefferson, however, later it became Ms. Daniels.  Our first day there we enter Ms. Daniels class room and the kids were wild.  Ms. Daniels is trying to settle the kids down, but to no avail.  So, she ignores them and asks us present!  So, in shock (neither Jesse or I expected to have to present Botball to the entire class of like 25 students) I pull out my laptop and get a video ready.  I connect it to the projector and start the Botball movie.  All of a sudden the room goes silent!  I could hardly believe it.  The movie is only 3 minutes and as soon as its over it is back to chaos.  Everyone is asking questions and wants to know why we are here, what other things can robots do, can we make robots that play football, etc…  So, we try and answer some of the questions and we finally ask Ms. Daniels where the Botball team is and where we are to meet.  Because there was no way could teach all 25 students robotics.

 

 

 

Some thoughts on MRL

MRL = Multirobot Learning

A quote from Multiagent Learning in Large Anonymous Games (Kash, Ian A et al.)

“With more learners, the noise introduced into pay- offs by exploration and mistakes becomes more consistent. Second, having more information typically improves performance. Publicly available statistics about the observed behavior of agents can allow an agent to learn effectively while making fewer local observations.”

So, for auctioning task would it be wise to be continuously auctioning off tasks to the robots even though they won’t actually be doing them so that they can learn how to bid more accurately in the different situations?  Use Dyna architecture.