People and bike counter (for use on a bike path)

[Liamlambchop]

Hello all,

My job is to source or make a durable, inexpensive counter which will be used to count the number of people who use an outdoor bike path. The counter will need to be able to count both people and bikes. I am thinking about using an infra-red beam which when broken by the upper body of a person would cause a count to occur (e.g 1st break in beam = 1 person/bike, 2nd break in beam = 2 people/bikes, 3rd break in beam = 3 people/bikes, etc). I then need to have the data from the counter sent to a computer so that the boys in the office can know how many people are using their bike path in real time. I am using a GSM modem to have this data sent to them.

So what I need is:
1. A counter. This can be bought, or made by me. Do you know of any commercial counters which would be durable enough to be used in rain, hail, and shine? Alternatively, are you aware of any instructables, circuit diagrams, DIY projects which I could use as a guide to make my counter.
2. A way to have my counter 'talk' with my GSM modem so that it can send the data from the counter to the guys in the office.
3. An end product which is cheap, easy to construct, and durable.

Wach'all think? Any input would be so appreciated!

 

[(*steve*)]

One obvious problem is that people walking in a group may trigger the device less than the number of people, and other circumstances may well count more events than there are people.

It's tricky, people are not regular and they don't travel in defined lanes

 

[Liamlambchop]

Yes well im banking on the fact that it's a bike path, so it'd be unlikely that there will be a lot of congested traffic.

Perhaps I should make everyone on the bike path walk in two straight lines, just like I had to do in primary/elementary school

 

[(*steve*)]

I travel along a bike path for about 5 km twice a day. I notice some behaviour that may be useful for you.

The first is that bikes tend to travel in single file unless the path is very wide. Picking a more narrow and/or closely fenced section of cycle path will pretty much guarantee this.

The second is that couples will tend to walk side by side. This will be a problem, but one which may be avoided by having the beam pass obliquely across the path.

I have noticed that where the cycle path makes a tight bend (or where there are bollards), people tend to fall into single file as they walk along that part. This happens even moreso when the visibility is obscured. In addition cyclists will rarely want to pass or overtake at these points so the counting of traffic may be more accurate here.

I have seen devices like this installed at shopping centres and I am sure that there must be ruggedised versions for use outside.

 

[Liamlambchop]

Thanks Steve! That is very useful information. I am not a regular user of bike paths so I would not have thought of these things.
Brilliant.

The idea has arisen of using buried pressure sensors instead of infrared sensors. This is to make the system more rugged and vandalism free. I don't think that vandalism is going to be an issue for my path, as it is in a rural area. Also, the path is already constructed, so in the interest of easy installation of my counter it would be preferable that I didn't have to dig the path up. What do y'all think?

 

[(*steve*)]

I was thinking about that. The problem I see is that cyclists represent a pair of pressure "spots" moving along, whilst people represent a discontinuous series of (kind of) single points.

A pressure pad too narrow may be stepped over, one too wide may capture several people. Because we all weigh different amounts, it's hard to use weight as method of counting how many people are on the sensor at once.

People walking together may be in lockstep, so even detecting foot falls may be inaccurate.

And you don't want to count cyclists as 2 people.

I couldn't figure out ways around these. Perhaps you could use a number of sensors, but it's starting to get hard.

Some form of artificial choke point may be a solution.

 

[Liamlambchop]

I think your idea of having the infrared sensor at an oblique angle will solve any issues of counting two people as one, and if it is put at chest height then it should be fine it count both pedestrians and bikes.
I think that the pressure sensor idea can be abandoned.
Also, your observations about when people walk in single file is valuable. That in itself is as good as a choke. If there are narrow bridges on my path (I will find out) they may be the perfect spot for the counter.
I really need to source a company which can provide me with a cheap counter, or source a DIY instructable which will tell me how to make one. I'm thinking of using an arduino microcontroller to be the counter, and just using a photodiode and a laser to be the sensor.

 

[(*steve*)]

IR lasers are dangerous. A focussed IR beam is safer. You should be able to find purpose designed devices to do this.

A quick Google suggests that things like this are available.