DCC++ to a new level

[MoPac]

Hmm? I am all for DCC and DCC sound, though to me DCC++/JMRI is still behind the curve. If someone can come up with a program that not only reads the locomotive but also knows where it is at all times on the track via electrical contact/current with the tracks and reading the decoder itself. Just like if you were giving it a command to increase the throttle or activate the horn. Doing this without using blocks and IR sensors. Yes, signals will still be used in conjunction with turnouts etc. Plus to step it up a notch and put a decoder in the caboose/EOT car so the computer can read it also. Regardless on how far apart the two signals are. the computer will understand that the distance between the two will/should stay constant unless operations will either add or subtract from the train. So whenever a train passes a signal the light will change accordingly. Also the computer should be able to detect breakaways if the caboose or EOT car suddenly stops and the locomotives are still going. If that happens this will stop the train itself and all other trains on that line or the whole layout. By preventing an pileup. Just a thought, I do have a more in-depth ideas on this. It is fairly simple, plus it would be based off of the old current detectors.

Any thoughts?

[SirTainly]

How would the computer know where the decoder was? DCC works via broadcast so it just sends out commands and whichever decoder recognises the id of the broadcast responds. You'd need to have the each decoder broadcasting a location back to the control station somehow. Unless you can have something above the layout simulating a kind of GPS, I'm not sure how you could make this happen.

[Rivet Miscounter]

The current detectors you refer to require blocks.   I don't believe there is anything "fairly simple" about what you are proposing.

Reading about LiDAR sensors in the new iPhone over the past few days, (hoping I can use it to survey my backyard, but mixed feedback about that) that is about the only technology I can think of that might be able to be used, but it needs a sensor that I'm not sure can be easily disguised in a locomotive.  Plus it would probably add enormous cost to the system.

[conrad]

Micro cameras and image recognition is another idea.

But, here's a slight digression:

Work is being done to upgrade/rewrite DCC++.  The new version is called DCC++ EX, see https://dcc-ex.com/
I have not used it yet as one to the lead developers recently had a KickStarter project failure (see the
thread "FireBox: A Tiny DCC Train Controller").
Check out the site but use at your own risk.

Conrad

[dem34]

What you are essentially asking for is

1. A system wherein the base station is generating a layout map based solely on its raw signal/output.

2. With that system existing, also somehow having a decoder function that transmits location data that corresponds with the map generated by the base station.

3. Having a fairly robust built in algorithm that functions like the AI in a Train Simulator.

The inherent problem is that while on point 3 there are a few programs through Digitrax and JMRI that do that now with a bit less flexibility, points 1 and 2 would certainly require a change in the hardware of both the decoders found in rolling stock and the base stations. Along with my doubts that even if something similar to point 1 was made that it would be accurate enough for me to trust it over just having IR/light sensor blocks.

[Maletrain]

The OP seems to be saying that he wants a location system "based on the old current detectors".  So, he would be able to tell which block had current, but not where in that block the current was being drawn.  If he wants to add a current detector to each drop in each block, he would still need to be able to figure out how much current through each drop means a locomotive is near that drop, because the bus and rails provide multiple current paths to any locomotives in the block.  And, there could be more than one train in the large blocks typically used by DCC layouts.  Even with sophisticated computer analysis of currents through drops, there is no way that the length of a train could be accurately checked with such a system, even with lighted caboose or end-of-train resistors.

Even with locos and caboose/EOT cars transponding their identities through the rails, there is no way to tell where those items are located on the tracks.

If we start thinking about some completely separate train location technology, like a mini GPS system, that would need to have really good positioning capability to distinguish which of two or more parallel tracks a train was occupying.  Overhead cameras with the type of computer assisted item image tracking that is being used for the aircraft-based crime-fighting areal surveillance system being tested in Baltimore would probably work, but I would not expect that to be anything like "cheap" or simple for a layout operator to set up.

This really says something about how much diverse information humans use to operate a layout smoothly, using eyes and ears and logic to integrate all we can can see and hear.  Trying to duplicate that with mechanical/electrical automation is like trying to build a prototype positive train control system in miniature. And might cost something like 1/160 what the prototype guys are spending to do that, now.

[MoPac]

Micro cameras and image recognition is another idea.

But, here's a slight digression:

Work is being done to upgrade/rewrite DCC++.  The new version is called DCC++ EX, see https://dcc-ex.com/
I have not used it yet as one to the lead developers recently had a KickStarter project failure (see the
thread "FireBox: A Tiny DCC Train Controller").
Check out the site but use at your own risk.

Conrad

I just read into this new DCC++EX this is on the line on what I was thinking of. Where the "Current reading, sensing and ACK detection settings in milliAmps instead of just pin readings" would play into this. The computer can read, sense and detect the current shift. Doing so it will not only know where exactly the train is but it would automatically understand the diagram of the track.  Yes, maybe a little high tech.

But imagine you are going to grab your beverage from the fridge or brew a pot of coffee. While letting three or more trains run on the track doing operations and a few just cutting circles. Then one of the trains EOT/caboose car and a few others along with it decided to breakaway. The computer will get the signal from the stopped caboose and then stop the train it was a part of and any other train that might be headed it's way. This is what I am digging into.

[MoPac]

How would the computer know where the decoder was? DCC works via broadcast so it just sends out commands and whichever decoder recognises the id of the broadcast responds. You'd need to have the each decoder broadcasting a location back to the control station somehow. Unless you can have something above the layout simulating a kind of GPS, I'm not sure how you could make this happen.

As I have mentioned before it would be done through the track itself not radio signal. Because not everyone is running their trains directly through radio/app control it has to relay the signal through the track itself right? Doing this the computer will know the location also. We are talking bout an smarter system here.

[peteski]

Not to pee on your parade, but depending on electrical current readings to that degree seems unrealistic (especially in smaller scales).  IMO, the wheel  contact with the track is not reliable enough for that).  Another sensing technology, such as RFID tags in each piece of rolling stock, and the readers embedded in the track woudl probably be a more realistic  solution.  Or barcodes on the bottom of each car and optical scanners in the track.

Or go full-monty and have overhead cameras in the ceiling tracking movements of the trains.

[MoPac]

The current detectors you refer to require blocks.   I don't believe there is anything "fairly simple" about what you are proposing.

Reading about LiDAR sensors in the new iPhone over the past few days, (hoping I can use it to survey my backyard, but mixed feedback about that) that is about the only technology I can think of that might be able to be used, but it needs a sensor that I'm not sure can be easily disguised in a locomotive.  Plus it would probably add enormous cost to the system.

No, it will not require blocks thats the old way. The cost will be more into software. The hardware is already there.

[MoPac]

[MoPac]

What you are essentially asking for is

1. A system wherein the base station is generating a layout map based solely on its raw signal/output.

2. With that system existing, also somehow having a decoder function that transmits location data that corresponds with the map generated by the base station.

3. Having a fairly robust built in algorithm that functions like the AI in a Train Simulator.

The inherent problem is that while on point 3 there are a few programs through Digitrax and JMRI that do that now with a bit less flexibility, points 1 and 2 would certainly require a change in the hardware of both the decoders found in rolling stock and the base stations. Along with my doubts that even if something similar to point 1 was made that it would be accurate enough for me to trust it over just having IR/light sensor blocks.

Now you are cooking!!!
With number 1 yes!
Number 2 I would also add that the decoder is transmitting through the rail via current is sensing the milliamperes and know when and where the locomotive is rolling by. Because you have to pair the Caboose also with the locomotive/s. Yes, the computer is reading any shift in the tracks layout. If the said value of the locomotive and caboose decided to expand while rolling down the track. The computer will get wind of this and stop the train and others that might be affected.
number 3 Yes, and yes AI algorithm altogether.

[Maletrain]

There is a lot of blithely stating that "the computer will know" being proposed without any description of the underlying physics or programming of how the computer is going to "know" where each locomotive and caboose are located on the track plan.  It is not as simple as reading currents, because the wiring is not necessarily following the actual track, and multiple drops makes multiple paths for currents.

So, how about addressing that?  And, please don't tell me to "go read [some reference] and I will understand".  If you understand it then you would be able to explain it, clearly, to the rest of us. 

There is already an RFID system available for model railroads - and it is not cheap.  And, it doesn't do what you say can be done in the way of real time train position tracking, at least not with any reasonable number of RFID detectors.

[MoPac]

[MoPac]