Low-cost bike detector could revolutionise city planning

Wouldn't it be better to use the sensor to do something else than count

Such as set traffic lights - in the same way as bus lane lights work (might want to wait for say 20 cycles before changing)

Trigger warning lights when a cycle lane crosses a side road,

Seems like they are aiming too low

One good thing tech this cheap can help with is providing evidence that new infrastructure is "working"

Rather than using this type of technology to count cyclists since the investment in cycling infrastructure is unlikely to materialise if count 10 or 100 cyclists going past a given point, can't it be used in a scheme to reward those that do and their employers. If the incentive to employers was good enough perhaps they could be persuaded to invest in infrastructure or at the very least allow Cycle to work scheme and a decent shower.

Quote:

Automatic counters are only accurate if set up on a cycle lane/track with no other kind of traffic intruding.

I'm not sure thats true... the website states:

"Knock’s tools for detecting bikes aren’t new. A magnetic panel the size of a grain of rice detects the distortion a bike creates in a magnetic field as it passes. An infrared camera measures the heat pattern of a human. The new bike counting device combines those observations and uses a speed calculation to guess whether the passer-by is in a car, on a bike or on foot."

No doubt there are *some* errors, but there is no perfect 100% accurate way of collecting large amounts of data like this. Even people with clipboard only collect a representative subsample. This is how statistics works.

100% agree with the JimKillock too. If people misinterpret or misuse the data, that is not an argument for collecting no data.

Software engineer developer writes article trying to sell his kit to make a fortune. Who would have thunk it.

Good quality monitoring is essential to get an idea of any changes occurring, casualty rates, and an idea of changes in cycle traffic due to various factors from weather through to changes in infrastructure (should they ever happen).

The problem is that it has to be done manually to be accurate (people sat by the side of the road, or nowadays a video set up which can be scrolled through quickly by a technician).

Automatic counters are only accurate if set up on a cycle lane/track with no other kind of traffic intruding. Since there are few locations where automatic counters can do this, they tend to be difficult to use. That's the reason why they aren't there, not because of a software issue.

That 'old style bike counter' in the picture is in Copenhagen, Denmark, not the Netherlands. Typical lazy Canadian journalism.

The other problem with monitoring is that it encourages this kind of thinking in some people:

Quote:

"Bridges, blimey, those things are expensive aren’t they? No, no, I can’t allow you to build a bridge. Not until you’ve proven that there is a real demand for this railway of yours. But you can build a bit of a bridge. A trial section of the bridge. We will monitor uptake and if it proves to be a hit, we will potentially allow construction of a bit more of the bridge, somewhere else on the river."

-- from https://waronthemotorist.wordpress.com/2011/02/05/fast-direct-uninterrup...

Strava Heatmaps can give you a very skewed demographic.

i.e. they'll show up plenty of "serious cyclists" who run Strava, but are likely to miss things like kids cycling to school, folk out doing some shopping, etc

Ultimately the bigger problem with any monitoring is that it shows where people are cycling but doesn't show where they would LIKE to be cycling.

But having some numbers is always useful - as long as they are interpreted with care.

Surely using Strava heatmaps would be cheaper

My local authority likes to justify providing no cycle infrastructure on the basis that almost nobody cycles.

With UK cycling numbers so low, I can easily imagine cycle counters being used to justify less, rather than more investment.