I am a third-year electrical engineering student and am working on an intelligent autonomous robot in my summer vacations.

The robot I am trying to make is supposed to be used in rescue operations. The information I would know is the position of the person (the coordinates of the person in a JSON file that can be changed anytime except during the challenge) to be rescued from a building on fire. I would also know the rooms of the building from a map, but I don't know where the robot may be placed inside the building to start the rescue operation.

That means I have to localise the robot placed at an unknown position in a known environment, and then the robot can plan its path to the person who has to be rescued. I can use gyroscope, accelerometer, magnetometer and ultrasonic sensors to do the localising job. I cannot use a GPS module or a camera for this purpose.

The object to be rescued (whose location is known in terms of coordinates & can be changed anytime) is surrounded by walls from 3 sides. Hence, adding more walls in this map.

According to my research particle filter is the best method used for localization of robot. But how can I deal with the landmarks (walls) that are fixed as shown in the map image and that are variable depending on the location of the object to be rescued being provided in the JSON file?

I can do the path planning from a known position to the target position, but I'm not sure how to determine the starting position.

More about JSON file: (1) json file containing the coordinates of the object to be rescued can change. (2) it won't change during the challenge. (3) json file will be provided to me in an SD card that my robot has to read. I have successfully written the code that will allow the robot to read the json file and hence the coordinates of the object to be rescued.

Here is the map of the building which is known to me.

enter image description here

  • 2
    $\begingroup$ Is this part of some challenge? I'm sure we had pretty much the same question on here a while ago. $\endgroup$ Aug 4 '16 at 13:57
  • 1
    $\begingroup$ Please at least link to the official problem so that it's clear what you ask for. $\endgroup$
    – FooTheBar
    Aug 4 '16 at 14:57
  • $\begingroup$ Okay, I made some edits and hopefully this makes your question more clear. Regarding the "best software" in which to implement something - generally the best software is the one you know how to use. It's generally an arbitrary/personal preference decision. It seems that your only question is how to incorporate particle filters into a localization algorithm. Is this correct? $\endgroup$
    – Chuck
    Aug 4 '16 at 15:13
  • $\begingroup$ Possible duplicate of Localising a robot placed at an unknown position in a known environment $\endgroup$ Aug 4 '16 at 20:39
  • 2
    $\begingroup$ @Rabia Khalid please always use an @ in front of names in comments so people get notified. Please clarify: are the walls known or not? What do you mean by "the json file can change"? Can it change during one "run" of the challenge? What exactly is in that json file? I find your use of the words "known" and "change" a little hard to understand. This would all be a lot easier if you simply linked to the official description. $\endgroup$ Aug 5 '16 at 7:42

Particle filter

According to the OP a robot with at least a distance sensor is available and a map too. That's a nice starting point for developing a hypothesis tracker aka particle filter. At first a game engine is needed which simulate the map and the position of particles. The game-engine calculates the expected sensor-information from the distance-sensor. E.g. the particle is on left top in the map and direction is "north" and the game-engine says "distance=10 cm". This is done for every random particle in the map.

Now comes the nice part: a particle will be deleted which has the greatest difference to the real measurement. That means, if the robot has a real distance of 5 cm on the sensor, than a particle with a distance of 200 cm can not be the right one. After deleting the particle the robot will be moved ahead, and the particles also. Again, the worst particle has to be deleted.

From a theoretical point of view, this is called a hypothesis elimination. It means, that the algorithm deletes the particle with the lowest probability to be the right answer. After a while a particle cloud with the real position of the robot is remaining.


Fixed landmarks in the map (like a door which is marked with a red line on the ground) can be easily integrated in the particle filter because in the map are particles who fulfill the conditions and others who don't. E.g. if the real robot detect a door, all particle in the game-engine who not detect the door are wrong. A more complicated task are variable landmarks which can be anywhere in the map. These obstacle change the map itself and create an extended problem called SLAM (=map building). Algorithms like FastSlAM and Hyperparticle (every particle has its own map) build a map and localize together.

  • $\begingroup$ Compliment to Chuck. You edited my answer in under 1 minute. Are you a chatbot? $\endgroup$ Aug 4 '16 at 19:02
  • $\begingroup$ No, I just saw the main page indicate a question had changed. I generally check the site every hour or so at work, as a break for what I'm currently working on. I seem to check it about the time people make changes. $\endgroup$
    – Chuck
    Aug 4 '16 at 19:20
  • 1
    $\begingroup$ "a game engine is needed" I find that term a bit misleading. It's a simulation, why not call it like that? Telling somebody to look for a "game engine" they will likely look at very wrong places as I doubt that anything that calls itself a "game engine" includes a particle filter simulator. $\endgroup$ Aug 4 '16 at 20:35
  • 1
    $\begingroup$ All we know right now is that @Chuck has passed the Turing Test. $\endgroup$
    – Ian
    Aug 5 '16 at 18:16

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.