Car Follower/Tracker

[eugenesiew]

Hi All,

I came up with an idea but i need some feedbacks from you all concerning its possibilities in implementing it. I want to make a car follower. Example, i place a chip on a lead car and another car follower is to follow the lead car. The car follower is to maintain a specific distance from the lead car. Meaning that if the lead car accelerates, the car follower should also accelerate... I understand that this uses fuzzy logic(do correct me if i am wrong or if you have better alternatives) concept and i am currently trying to learn it(do send me any revelant materials concerning fuzzy logic).

For my project, the lead car will be just an ordinary toy car in which i will be attaching a string to pull it. The car follower would be the actual thing that i will be designing....

The question is:

1) Is there such a device to be place on the lead car so that the car follower can detect it and aim toward it? if so, where do i get it?

2) Most of my seniors using the H-Bridge motor thingy, but some people suggested to use stepper motor instead... (stepper motor is expensive and i plan to get it from my spoiled canon LBP1120)? hows that?

3) The car follower should avoid any obstacle that come in the way, that could be done with IR sensors i suppose right? or is there any more efficient alternatives?

4) I thought of using PIC as the microcontroller (as i shall be undergoing a PIC training course) to be brain in sensing the distance and velocity of the lead car and do the necessary adjustment for the car follower speed, can PIC do the job? meaning the PIC has to be able to detect the lead car and on the same time avoid obstancles in the path...

5) I have only 4 months to implement it... do you think it's sufficient enough for a diploma student?

Please feel free to give any ideas concerning my idea. Thanks a lot in advance!

Eugene

 

[TrevorP]

Well ok you're going to need some sort of thing to be tracking on the car...and then two detectors spaced out so that you can determine to turn more to the left or right.

This is actually a pretty interesting idea...I know at my university they do a robot building competition in the summer after third year. (Not the senior year project.) But it's usually only a obstacle avoiding racer...actually probably quite similar to what you are doing.

 

[civictypeR]

Of course you have more than enough time to finish your project. Sounds intersting , It seems you may need to use GPS system.

 

[johnnitwenty]

it's so easy,just do it as a train.
it's difficult to finish the system. think as the car(head or follower) ,you will get more inspiration.

 

[mazahar]

I need car tracking system design.

 

[tcmtech]

I have seen demonstration videos of such concepts done by big research and development companies.
You will need about $15000 per vehicle in equipment to get it to work right.
Every car has to have the throttle, brakes, and steering integrated into a central processor that has to have a continuous link with every vehicle in front and behind it.
That way when one car turns off the road every one behind it does not follow it.
Even then its still not all that practical.

So if you have a million dollars or so to spend an a few good buddies in automotive modification and wireless computer networking that dont really have much going on for the next few years take a shot at it!

 

[shimniok]

I think he is talking about something miniature (pulling car by string) ? That would greatly simplify the cost/scope/time.

Not that I have done anything remotely like this, but ... I have seen numerous different approaches to following...

I understand that some camera systems can help with tracking of a color (CMUcam). Or perhaps use some kind of beacon (IR pair) ? Perhaps you can use a different system to detect and maintain a constant distance?

I think 4 months is doable depending on scope and how much you know, how much time you have to work on it, etc.

Projects like this are easier, at least in my experience, when you can document what you want it to do (requirements), then worry about how to get it to do those things later (architecture & design).

Also easier if you can sort out the requirements into logical functions -- basically breaking down the problem into smaller problems that you can independently solve more easily. E.g., the "car" has to accelerate, decelerate, and steer. The functional chunks are "speed control" and "steering". Another function is "track and steer" -- detecting the other car and what direction it is in so you can steer correctly. Another function is "maintain distance" -- measuring distance and altering speed accordingly.

Hope this helps

Michael

 

[Wond3rboy]

I have seen a numebr of 'car tracking' projects here. As a general direction use a GPS,SIM(atleast thats what we use here for 2.5G-2.75G networks). Use a microcontroller with the sim, a lot of people have done it(usually as a train tracking system) but it applies to a car in the same principle.Its a nice project and has been commercialized here(big $$).