Must watch movie

We had a great weekend of Death Note last week, this week highlight, Ketika Cinta Bertasbih 1 & 2. A great movie..10/5!!!!

Advertisements

Alan’s comment

We had a fruitful discussion with Alan Winfield on how we can design our granuloma formation algorithm with the robots. Questions that I have in mind are now answered, including on how can I attract the functional robots to move to the faulty robots. Thanks Jon for arranging this meeting!!!! It is priceless to meet Alan, he offered us a place in Bristol Lab for a week or two if we plan to design the e-puck later because they have the equipment. He even will send a thesis of his student so imitate the Nembrini’s Work in a simpler way.

Alan’s student Idea (yet to be published)

In a simpler way than Nembrini, he does not look at the neighbour of the robots, rather than checking on the collision status. If there is a collision, means they are in a right direction because they are together with the swarm of robots. However, if the collision detected from the t(last) = 0, means they need to turn back and move until in t, collision detected = 1.This idea get the same result as Nembrini’s beta algorithm which, the aggregation of swarm is still maintain and they can move together. It’s simpler and more efficient as mentioned by Alan. The student’s viva is up next week and I think after that we can get the copy of his thesis for reference.

In term of faulty robots and energy sharing, Alan gave his opinion  on how to design the algorithm. In our previous experiment, we use a beacon to attract the robots’ movement. Here, he mentioned the faulty robots now act as a beacon. Each robot need to have a LED and LED sensor. If the robot is in functional, the LED signal is 0. If it is in a faulty state than the LED signal is 1, which informing the nearest possible robots to move towards it. Once enough robots are there, Alan said, it is best if the LED signal is disable so that the other robots keep moving rather than waiting for the recharging task is finished so that once the recharged, the robots will then move forward to the first cluster of robots that have already move in front.  Keeping that in mind, we are now getting closer in designing the simulation of our swarm of robots.

Algorithm:

So, this week I need to write down my algorithm to be included in the conference paper that I am preparing. Things that need to be updated this week:

Swarm aggregation task Achieved:

-reproduce results from Nembrini

-added energy to the algorithm

-added analysis on charger to the algorithm

Granuloma formation algorithm In progress:

-change from Nembrini to collision detection algorithm

-added energy

-add LED signal

-isolate them first

Once isolated:

-recharge and move

– current status, lookin at on LED signal in player/stage

– still writing, need to update the bibtex file for the conference…better do it now or else..the paper will never be submitted on time!!!