Help: Solution for Detecting the No of persons sitting on a seat

Iam working on a project that will detect the number of persons seated on the bike, and if the number of persons exceed 2 then it will immobilize the engine, The Challenge is to detect the number of perons seated. you can see the link for reference of the Bike seat http://www.heromotocorp.com/en-in/the-bike/two-wheeler-motorcycles.html . By the Bike i Mean a Two Seated Motor Cycle.  if you have any better solution apart of detecting by the persons sitting on SEAT kindly enrich with that.

The government regulation is 2 persons but sometimes people go with morethan 2 posing risk this is the project.. i think i have elaborated all.;)

This sounds like an EE senior design project to me, am I correct?

Edit: Maybe too advanced for a senior design, so a Master's Thesis?

Iam a Masters Student though.. but its not for thesis just a project that i have to get done..

Iam a Masters Student though.. but its not for thesis just a project that i have to get done..

In this case, we should approach it differently. Tell us some ideas you've had about detecting the extra persons, and we'll make recommendations on how you could proceed with them.

Michael Wylie wrote:

 

In this case, we should approach it differently. Tell us some ideas you've had about detecting the extra persons, and we'll make recommendations on how you could proceed with them.

I agree; we shouldn't make his homework. There are some nice technologies available that could be useful, but Eeshan, please tell us first how far you've come / what ideas you came up with by yourself.

First things First:: Please dont worry about you  doing my homework i have came to this forum only after getting stalled. Moreover iam a Research Student, School times have passed things arent proving that lucky now.

1, I tried to use Weight Sensors But no relief the human weights vary largely from 40kg to 130 kg/person so i cant have a reference weight. Moreover if i set a maximum limit of 150 kg for average for 2 persons the same gets violated by  a single person carrying 2 kids.

2.I thought of using sensors on Foot rest pedals there to some times while riding off limits the 3rd person leaves his legs hanging as there is no Foot rest pedal to rest on as the bike is designed for 2 only.

3. I tried using bubble switched but got limited as each switch took a i/o pin. Then i tried to use the matrix type press detection as we use in Keyboard in that the problem encountered was when a switch gets pressed the port becomes High and remains high making the entire row and column in high state while as in keyboard the key bounces back.<<New Ideas&techniques are Welcomed on this>>

4. Now trying to use the principle of transformer(emf). I will put a layer of conductive solid surface which will be having a definite charge as base  and on top  i will put strips of another conductive element they will be separated by a dielectric between them. As the person sits the strip will be sandwiched resulting in more emf on the wire.. so i will put many wires and detect the change in the seat load and by that i can get pressure points too and detect the persons seated..<suggestions welcomed>

5. The aim of the project is to be a cheap yet effective solution, i could use a dozen high tech sensors and controllers but lets not forget it to see it from market perspective.

APPRECIATED!!

I think you should have disclosed your work thus far initially, but that's just my opinion; take it as such.

I think I have to agree with Victor, the camera and image processing will be the best result. You might even need two cameras, but the hard part will be where to mount the cameras. I don't have experience in this area, but I've seen what cameras and image processing can do. That's my two cents.

Eeschan from what I've seen when the Bikes are over loaded they have the whole family and often some livestock attached.

As you can imagine not all are on the seat so maybe that's not the place for the sensor have you considered on the rear/Front suspension.

In fact use the deformation of the suspension as the sensor! Also measure the number of times the weight goes up by a reasonable amount not the actual amount

BTW So far around Delhi the record I found was 7 people  and shopping on a 50cc Motorbike!

Michel this was my first post in Element14 i didnt even know that i would get a response.  I know that Image Processing would do but in Two wheeler i don't have a definite place to mount making it unsuitable.

Sensing the suspensions may not be a good idea because this idea is the same as the weight sensors. I mean to say the same limitations.

I think the weight one could still work. It's not the maximum you need to worry about. If you place one long strip of strain gauges down the center of the seat you could look at the strain distribution along the seat. Two strain peaks for two people, three peaks for three people, etc...

Great post Eeshan, thanks for clarifying your thoughts.

Here are some ideas that might be interesting:

- Tekscan makes these foils: http://www.tekscan.com/body-pressure-measurement They're probably too expensive and need too much callibration for your 'product', but might be interesting for R&D.

- Have a look at this: http://www.takktile.com/product:takkarray These people make force sensors by encapsulating atmospheric pressure sensors. low cost, open source.... Might be nice for you to make a prototype. You can use these for prototyping:http://www.takktile.com/product:takkstrip . I believe most of their products are open source.  With a grid of these sensors you might be able to do a pressure mapping

Your idea 3 could be feasible when you take one switch per pin, but multiplex the switches to the microcontroller. So don't make an array, but use multiplexers / shift registers to read out the state of each switch.

Your idea 4 looks like a resistive touch screen. You might want to experiment with those, I don't know how bendable those stick-on touchscreens are.

Have you looked at capacitive sensing? You could try to get a grid of capacitive sensors. To get started easily you could use these chips, again combined with shift registers to read in the data: http://uk.farnell.com/atmel/at42qt1010-tshr/sensor-touch-cap-1ch-6sot-23/dp/1841593. Rain could be an issue, but when this would become a product you'd go for some other capacitive touch solution with more control over oscillation frequency / thresholds.

Greetings,

Victor