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

Hi Bashir

Here is an idea. I am going to assume that only one person at a time can add their body to the bike. We are going to put a pressure sensor, like an bladder into the seat and connect it to a pressure sensor. We are not going to be concerned with specific pressures as people sit on the seat but rather we will detect pressure plateaus. We will use a micro processor to count the pressure plateaus. For example the first person sits on the seat and a max pressure, that is stable for a second or two, is detected. The next person sits on the seat and a second pressure plateau is detected, which is also stable for a second or two. Finally if a third or fourth pressure plateau is detected then the microprocessor disengages the ignition. You would have to program the processor to test for any pressure that falls within a range plus or minus "x" Kpa for a second or two. If a second greater pressure is then added and stabalizes for a second or two we have our second person. Then if a third higher pressure is obtained we have our prevent ignition point. You will have to make sure that there is a fail safe so that once the bike is started and moving that the ignition shut down can not happen as the bike and people bounce over rough terain. Perhaps a timer that takes the shutdown out of the picture once the bike has been started and moving for 5 seconds. Good luck

John

@John Wiltrout : This case is already discussed as taking weight sensors and detecting the concentration points. Please elaborate the Bladder concept if it uses only a single sensor then it can work.

The bladder is an air or liquid filled bag that is under the upholstery of the seat. When someone sits anywhere on the seat the bag is squeezed which increases the pressure which in turn is sensed by a single pressure sensor. If a second person sits on the seat the bag is squeezed more increasing the pressure. If someone sits on someone else's lap the bag is squeezed more and the pressure increases. We do not have to worry about how much people weigh as we are not looking at any specific weights but only at stable pressure ranges that are above the unloaded pressure. When the microcomputer detects a pressure in the bladder that is above the base level it will count one person. When a second relatively stable pressure is detected it will count two people and if there is a third relatively stable higher pressure level we have counted our third person. This will not protect a situation where a mother with a child on her back gets on the bike as our sensor/microprocessor will see only the addition of one mass but if the mother gets on and then the child climbs into her lap it will detect the child. By counting stable pressure levels we can illiminate the complication of trying to estimate weights or compensate for the difference in boddy shapes and sizes. Even if people put some of their weight onto foot pegs or the ground, by detecting maximum pressure plateaus we can determine that only one person is being sensed by programming the correct code into the microprocessor. I hope this clarifies my idea.

John

John, that's exactly like I said. But instead using pressure I suggested measuring movement using a metal plate and induction. Both solutions need some form of calibration.

Elbert.

Hi Elbert,

I believe that you may be missing my point. The method I am suggesting would also work with your sensor but instead of looking at absolute values I want to look at the rate of chage of the sensor measurements. If we take 100 sensor readings a second and we compare each sensor reading to the average of the last 100 we will know when the mass has been put on the seat as we will see an increase in the numbers and then a leveling off of the numbers as the person comes to rest on the seat. We don't care if the person weighs 20 Kg or 80 Kg just that the readings have gone up and then stabalized. The secont person will cause the sensor reading to again increase to a average higher level and then stabalize at the higher value. The micro processor can keep track of these steps of increased mass on the sensor. When the count exceeds two processes of going up and stabalizing the micro processor can do what it needs to do. With this method no calibration or estimation of a persons average weight needs to be taken into account. The reason I chose the system with a pressure bladder in the seat is that my experience in more in pneumatics and I have worked with this type of system before. You are likely much better with variable inductances and could undoubtedly rough out a system using those sensors better than me. I believe my idea will work in a laboratory environment. Whether it can be tweeked to work in the real world is another question. If I was doing the project I probably would not mess with cutting the ignition but rather have the sensor illuminate a warning sign about the danger of more than 2 riders and let the driver make his own decision.

John

If you are looking for movement, picking up a bag or anything would cause your "stabilizing counter" to increase therefore someone picking up two bags would appear as three people.

If you are looking for movement, picking up a bag or anything would cause your "stabilizing counter" to increase therefore someone picking up two bags would appear as three people.

This has grown into quite a discussion. In my opinion, measurement of change in sensor value is by far the best solution. Suggested by many in the discussion primarily as John Wiltrout suggested. Just an additional tweak, How about a using an ultrasonic distance sensor mounted on the bottom of the bike centered between the wheels. Simple, easy to detect difference in weight (distance to the ground) easy interface to a microprocessor, one sensor with no calibration, only looking for stabilization. One demerit I can think of is if the kickstand was down when the passengers mounted, then when the weight was taken off the kickstand the sensor would register another passenger. Possibly corrected by a switch on the kickstand to prevent evaluation for the next x milliseconds?

Bill

RE: Kas

Maybe set a minimum increase. That is recognize only a change of xx representing say more than 50 Lbs?+

Hi Kas,

You make an excellent point. these are some of the hard side effects that will have to be solved if the idea moves out of the lab and into the street. Part of the fun of a discussion like this is seeing if we can come up with a solution. How would you engineer a fix for the cyclist that wants to load two bags of groceries onto his bike before he heads home?

John

Hi Bill,

Your idea about the ultrasonic sensor is a good one. It would eliminate the problem with my bladder in the seat idea not detecting someone riding on the gas tank or the handlebars.  Another problem that comes to my mind: I am not a motor bike expert but don't a lot of them lack electrical systems that could maintain the circuitry needed to keep the electronics alert and ready for the rider to start loading up the bike if it was left unused for a couple of days. It seems to me that they are generally magneto powered with a very small battery. Anyway this problem will need a better man than me to put it on the road. At least we gave Eeshan some things to think about.

John

Bill-SD You are missing the point,, measurement of weight wont do because i can load three persons of 50 kg and two persons of 75kg same deflection...

Hi Bill,

The ultrasound sensor will work for some time, but then the accumulated mud and dirt on the underside of the motor will block any ultrasound transmission. Depending on the user it will fail within a few weeks after the bike was bought...

look at John #32 and couple it with the sensor suggested. Not looking for weight but a stable peak. As for the mud I'll bet you can find a protected place or build one easily. ( Is this going to be a Dirt bike)?  If it is an off the road situation then up comes a new set of obstacles. maybe use two of the sensors, one on front and back of the shocks,springs. You may have to go to a more sophisticated sensor such as induction or laser. You have a wealth of suggestions. Look at combining some of them. I'm pretty sure that some of the current microprocessors with very low current consumption coupled with distance or angular sensors could solve the problem. Don't look for an altogether complete answer but look at all the suggestions and contemplate new ideas derived from the ones already given.

Bill