Stats

Full 24 hour period
Date	Events	Shortest	Longest	Average	Down Time
2026-10-10	93	15	625	218	23.4%
2026-10-09	107	54	505	233	28.8%
2026-10-08	9	46	345	158	1.7%
2026-10-07	107	30	630	237	29.3%
2026-10-06	111	20	828	239	30.7%
2026-10-05	96	31	997	253	28.1%
2026-10-04	58	30	350	178	11.9%
2026-10-03	82	30	552	208	19.7%
2026-10-02	105	69	766	235	28.5%
2026-10-01	109	24	589	244	30.8%

Basic stats give per day seconds of crossing down and percentage of the day the barrier is down (subject to the mentioned limits)

Between 7 and 23 hours
Date	Events	Shortest	Longest	Average	Down Time
2026-10-10	79	15	625	216	27.9%
2026-10-09	87	54	499	228	32.4%
2026-10-08	0	-	-	-	-
2026-10-07	87	68	630	236	33.6%
2026-10-06	93	20	828	235	35.8%
2026-10-05	82	31	997	255	34.1%
2026-10-04	58	30	350	178	16.8%
2026-10-03	66	30	552	214	23.1%
2026-10-02	87	69	766	236	33.6%
2026-10-01	92	24	589	237	35.6%

Between 23 and 7 hours
Date	Events	Shortest	Longest	Average	Down Time
2026-10-10	14	134	430	226	9.8%
2026-10-09	23	135	505	255	18.1%
2026-10-08	9	46	345	158	4.4%
2026-10-07	24	30	460	246	18.2%
2026-10-06	23	103	530	247	17.6%
2026-10-05	18	140	526	250	13.9%
2026-10-04	0	-	-	-	-
2026-10-03	16	115	310	181	9.0%
2026-10-02	23	65	425	225	15.9%
2026-10-01	23	85	530	263	18.6%

I'm open to ideas on other useful stuff - histogram plots of operation against time of day might be interesting

Anyone interested in turning crossing up and down times into some interesting stats - feel free to get in touch

If you want to play with raw data, it's here. Each crossing operation gets a row entry, the first column is when a signal goes green (crossing down), the second is both signals red again (crossing up). Time is Unix time stamp (seconds since 1970)