@ stevepresley -
It is terrestrial, we have recently changed the model to eliminate the master/slave architecture, and are implementing peer to peer communication between autonomous objects. Each object will have continuous real time communications with all other objects. The goal is to define a series of response algorithms based upon the information being received such that the objects will behave in a desired way given the random locations of all the objects that are within a given radius of a know location. By the way I found this article of some funded MIT development that shows hope for indoor close proximity awareness at low cost. Thought everyone might have an interest. My experience with this group tells me you already are aware. :-). http://cricket.csail.mit.edu/
Awesome, thanks for the info on your project.
I wasnât aware specifically of the Cricket project, but it is very similar indeed to what has been described in this thread. Many good links off of that site as well that I will be reading a couple of times in the next few weeks.
@ stevepresley -
Do you think GHI would have an interest in using Cricket open source to engineer a module? It would provide a solution to a wide range of problems.
keep in mind that gadgeteer is not only GHI. The community has made much and more on the way as well http://www.tinyclr.com/forum/topic?id=6208
I donât know that GHI has the bandwidth specifically, but I would imagine a few of us on this thread would want to engineer something, whether Cricket based or not. Iâm certainly willing to contribute my work with GPS, Ultrasonic and IR. I have RFID and QR code experience as well, so working that in as markers could be something to investigate. I have some Bluetooth experience, but no where near the expertise with network level transmission protocols that you and others have chimed in on, mainly from an application transport/development perspective.
@ Bob - you still wonât have the speed to do a cricket design in .net
Hereâs a TED video about flying swarm bots that have solved many of the techniques we have discussed in this thread:
http://www.ted.com/talks/vijay_kumar_robots_that_fly_and_cooperate.html
@ stevepresley -
Awesome!
@ Justin -
It would be built as an intelligent module that would provide answers NETMF.
Can the intelligent devices see each other?
There are several youtube videos of devices using RF to triangulate position of something in a small space using signal strength. The disadvantage is that you need to offload the computation of where something is due to the complexity of these algorithms.
As mentioned elsewhere; if there are obstacles this will not be accurate. Im working on a similar problem using image recognition where I have the advantage of fixed rotation so the problem becomes easier to solve.
Ultra sonic range finders are accurate and do work; but it doesnt tell you where the other object is. The same problem arises with any form of range detection.
Seems like they simply splatter a bunch of nodes everywhere, and detect where the âtracked objectâ is based on which nodes can see it. That wouldnât work for distances in the 20 feet range.