It was possible a note_kinematic_activity() call could increase last_kin_move_time, but _process_moves() could reset it. Fix by making sure _process_moves() only ever increases last_kin_move_time. Signed-off-by: Kevin O'Connor <kevin@koconnor.net>