Reply to thread

Why not just do this yourself instead of hiring a consultancy and bashing their design?

You don't want or need DC coupling because sunlight, oncoming headlights, thermal effects etc. will generate photocurrent which can saturate the amplifiers. Even with optical filtering.

I don't expect fog to attenuate IR very much at all. You can look at detecting cloudy days, the light spectrum, colour temperature, IR amount changes very little.

For ideas look at photoelectric smoke alarms, they use a multipath system to get a good air sample. I would also look at a differential system with two photodiodes to reliably cancel out drift.