I just finished a "game" session on my "Time Waster" switching layout. 5 switches and 9 sensor tracks in a 75" module. The end goal of building this layout is using LCS to provide automated shunting of cars from the starting positions to end positions with the classic Time Saver layout's restrictions on siding lengths. A dice roll is a good way to order the cars on the main line with another roll to determine destinations. There are practical problems to automation.
LCS signals the passing of locomotive or the (rare) sensor car. The presence of, and number, of cars coupled to the front and rear of the loco is unknown so the controlling software has to maintain the state of trains on the layout. Lionel needs to provide additional sensor car and train detection options. How about a kit to retrofit cabeese or other cars in our rosters to add input to sensor tracks? I'm sure Ryan can comment on how simple this can be if the LCRU receiver and logic in the sensor car is replaced with a simpler configuration mechanism.
The 1 to 98 address space of TMCC1 and TMCC protocols can quickly fill up with engine sensor IDs, but TMCC 0xFE and TMCC2 0XF8, 0xF9, and 0xFB commands can be extended to support a multitude of sensors (what was planned for the 0xFA address space?).
How about a "Train on Track" detection using insulated rails that feeds into a STM2 clone that transmits a "train entered" or "train exited" for a number of attached detection tracks and can be queried by software to learn the state of the detection track. (An insulated rail detection track that feeds a DC polarity signal (occupied | empty) to the STM2 is very feasible - I leave that board design to the H/W gurus).
With a shunting layout, command latency is an issue - the sidings are not long enough to allow slack in stopping the switching engine before cars are pushed off the end. In particular, Lionel's LCS App's suppressing control input for 3 seconds while it displays sensor track information is a big design miss. Wrecks await. I'm hoping that a Base3-related update to the App will fix this. The latency issue is a problem with all automation software as our track plans are compressed representation of a prototype and leave little slack for managing train speeds. Better detection of train position is essential.
Don't get me started on recalcitrant coupler opening and closing. Grrrr.
Automation with existing LCS products is going to be a challenge. Lionel has opportunities here.
