Live data would be great, but I'd be interested even in historic data. Not like 1880s but a timeseries of 15-minute granularity of train positions for last year

Live data would be great, but I'd be interested even in historic data. Not like 1880s but a timeseries of 15-minute granularity of train positions for last year. It'd be interesting to see what the bottlenecks were and what other routing possibilities there could have been.