I definitely noticed the jerk issue at Blackfriars junction on Thameslink, when trains stack up. The Thameslink ATO seems to cope well with the core section where it doesn't get too close to the end of an MA, but waiting for a queue of trains into blackfriars seems to regress into the full speed/full brake mode of operation. Perhaps because it's late on the timetable? A human will know when the junction has cleared for the train vs when the train needs to slowly move to next in line. With the Thameslink programme serving as a pathfinder for ATO over ETCS, I wonder if these lessons learned made it into the spec or implementations.
I definitely noticed the jerk issue at Blackfriars junction on Thameslink, when trains stack up. The Thameslink ATO seems to cope well with the core section where it doesn't get too close to the end of an MA, but waiting for a queue of trains into blackfriars seems to regress into the full speed/full brake mode of operation. Perhaps because it's late on the timetable? A human will know when the junction has cleared for the train vs when the train needs to slowly move to next in line. With the Thameslink programme serving as a pathfinder for ATO over ETCS, I wonder if these lessons learned made it into the spec or implementations.