That LTDA video makes me more determined to set up my own unbiased* study.
Pick a busy rush hour junction, rig up a cheap camera or my Raspberry Pi
Then collect it & review the results to see how many cyclists jumped the lights vs cars and work out the %'s and see which group is the worst offender.
*If the results are not in the cyclists favour, I'll just pick out all the cabs running the lights/stopping in the ASLs and use that in the final edit.
That LTDA video makes me more determined to set up my own unbiased* study.
Pick a busy rush hour junction, rig up a cheap camera or my Raspberry Pi
Then collect it & review the results to see how many cyclists jumped the lights vs cars and work out the %'s and see which group is the worst offender.
*If the results are not in the cyclists favour, I'll just pick out all the cabs running the lights/stopping in the ASLs and use that in the final edit.