Beyond DCC

[C855B]

This I thought worthy of its own thread:

... DCC is, at best, a transition technology--and one that's becoming relatively ever more expensive.

Why does a DCC base station cost more than a low-end PC? Why does a wireless handheld throttle cost more than a dual-core tablet? Because DCC is a technological cul-de-sac. The entire model railroad industry probably has a smaller budget than what's spent on wireless internet R&D each year. Bluetooth receivers fit inside a USB port. Google's Chromecast fits a wireless HD streaming-media player in a form factor comparable to some sound decoders.

Power through the rails, control through the air. Using Internet-standard technologies, please, not a 1980s technology that's first cousin to the Lionel Railscope. The choice between DC and DCC is like the choice between riding a horse or an ox while everyone else drives nuclear-powered DeLoreans. 

Damn straight. The chipsets exist right now to put WiFi on an N-scale loco decoder board. Control your MRR with direct comm between your WiFi-enabled smart phone and your locos. Large layout? Wireless router boosting the comm. Or let JMRI do some of the heavy lifting if you want. Doesn't matter.

Bottom line is that DCC is a 1980s technology relying on proprietary protocols operating on proprietary hardware. We are so beyond that now with other consumer products. The better MRR mousetrap is out there, waiting for somebody to integrate into a product.

[peteski]

Sure, there is always something new and better coming along. Especially in electronics/computer technology. Just look at Apple (or any other computer or cell smart phone manufacturer).

But majority of model railroaders are very set in their ways. There are new layouts still being built with DC control.    I would venture a guess that digital train control (that also includes systems like Rail Lynx, AirWire and Selectrix) probably make up less than 50% of all the layouts worldwide. DO you think that if some new-fangled system came along, it would quickly find wide acceptance?  Me thinks not.

DCC is a bit cumbersome, I admit.  Especially the programming part.  But it has plenty of capabilities.  What other capabilities do you need in N scale models which would require a completely new control system?  After all trains don't fly or drive on roads. They follow metal rails. They just need to have controllable motor, all sorts of lights and sound.  No other bells and whistles are needed.  Or am I missing something?

[DKS]

Bottom line is that DCC is a 1980s technology relying on proprietary protocols operating on proprietary hardware. We are so beyond that now with other consumer products. The better MRR mousetrap is out there, waiting for somebody to integrate into a product.

And what, pray tell, are all of the modelers who have already invested heavily in DCC to do with all of their expensive hardware when the next bright shiny objects come along?

[conrail98]

Bottom line is that DCC is a 1980s technology relying on proprietary protocols operating on proprietary hardware. We are so beyond that now with other consumer products. The better MRR mousetrap is out there, waiting for somebody to integrate into a product.

Really? I could've sworn that the DCC spec is open and found on the NMRA website. Oh yeah, here it is

http://www.nmra.org/standards/DCC/standards_rps/DCCStds.html

And I thought there was an entire organization devoted to its open source ability to roll your own effort, oh yeah, they're here http://www.opendcc.de/index_e.html.

I've been keeping my eye on this project and others. Ultimately, you're going to see DCC controllers/command stations on things like the Arduino very shortly. Bottom line is you'll never see anyone go beyond DCC unless the NMRA makes a spec of it. Why? Because no manufacturer is going to stick their neck out and develop something brand new that is only used on their system. That was the whole point of DCC and its spec. See MTH for why going off on your own is a bad idea. We *may* get to have WiFi controllers on-board but I'll guarantee you they are running DCC signals to/from a central command station,

Phil

[Philip H]

And what, pray tell, are all of the modelers who have already invested heavily in DCC to do with all of their expensive hardware when the next bright shiny objects come along?

Buy the Digitrax and NCE upgrades, like we do now.   

[DKS]

No offense to Zox, but this argument pops up on a fairly regular basis on many forums, with little regard to the financial realities involved. The fact of the matter is that dual-core tablets, Bluetooth, et al, are part of a market that is many thousands of times larger than the MRR market; hence, the R&D to develop such stuff is driven by the billions spent by both people and business on these technologies. The MRR market does not have anywhere near the critical mass necessary to drive the R&D required to deliver equivalent technologies. This is not to say DCC is totally stagnant and that we'll never see change; it simply will not keep pace with bleeding edge gadgetry. Quite frankly, I'm somewhat surprised that the MRR market has such a strong technology component as it is.

[C855B]

Bluetooth won't work, comm range is too short and there are too many consumer products under-implementing the RF power support. It's got to be WiFi or maybe something like ZigBee. But I wouldn't advocate ZigBee because it is, as David sort of inferred, another bleeding-edge consumer tech, and is not yet well-supported, i.e., still at "early adopter" price points.

"Proprietary"... well, actually, that's not a good way of putting it. I mean that in the sense of requiring dedicated equipment and control structures that can't be used for anything else, or sourced by anybody else other than a DCC product maker. Yes, it's an open standard and I have the docs, too... and the learning curve is pretty steep, especially when you're after bi-directional comm.

What is wrong with a system that uses consumer hardware and general-purpose OSes as the control system? Nothing that I can think of. JMRI proves that, but as a front-end to DCC it is a very clunky control chain and only adds to the cost of deployment, "cost" meaning setup and configuration time and effort, too.

...you'll never see anyone go beyond DCC unless the NMRA makes a spec of it. ...

The NMRA doesn't and won't have the staffing horsepower to deal with new-tech standards effectively. It took nearly a decade for them to publish DCC. If they had a paid staff of technologists and a true R&D budget it would be different, but relying on well-meaning and even very knowledgeable volunteers to advance the art takes me back to my ham radio days, ham radio now mostly an anachronism in the smartphone + internet era.

The advocacy for DCC that I hear here is that it is the status quo. However, after all this time there is still ongoing DC vs. DCC debate, so DCC is not by any means "entrenched". And I believe I have a solution that makes a bridge from DC block control more palatable. I am looking for technological reasons this shouldn't be pursued.

[DKS]

I am looking for technological reasons this shouldn't be pursued.

There's nothing I know of that would prevent anyone from adopting any sort of existing (or even developing new) technology. The stopping points are the cost to research/bundle/market the systems, and the high probability of a low uptake (not enough early adopter modelers) to amortize that cost. You can blue sky all kinds of fantastic train control systems. But how will they be funded? Figuring this out is far more important than product design.

[conrail98]

Mike, I'll give you another reason on top of what DKS said. FCC regulations of products using WiFi. From all indications I've read in hobby magazines or heard in interviews on the various podcasts, this is the main reason you have not seen any direct use of all the items you are mentioning and why going through JMRI for these types of setups is probably going to be the de-facto standard moving forward,

Phil

[C855B]

Well, this isn't quite "blue sky". This is mostly moving the comm and control layers from a proprietary implementation to one more in line with available consumer and business support. I recently supervised deployments of a similar nature in intelligent transportation systems (ITS), where we were making the crossover from internal standards relying on a lot of copper and custom RF to IP-based, stack protocols. No, it wasn't free or particularly painless - some vendors were less flexible than others - but implementation costs were 1/10 of the systems it replaced, plus we gained compatibility with the public internet.

As far as marketing such a system, and gaining enough acceptance to sustain it? It would have to be in partnership with an existing DCC firm. My EE knowledge is too out of date to design at the board level, so I can't create a working prototype on my own, but a good coop arrangement with some funding participation on my end might be enough to bootstrap something.

... FCC regulations ...

Now that is a good point. I cannot completely recall if the FCC certification for WiFi chipsets comes at the chipset level as long as there is an integrated antenna, or if the "finished component" has to meet compliance. I remember doing some work in this area about 8 years ago, and if it was a fully-integrated module with antenna that you just laid on your board, the compliance carried forward. A way around this might be to contract or otherwise encourage consumer production of a WiFi dongle on the mini-USB standard, and supply a mini-USB socket on your board. There is a WiFi dongle being sold right now for $10 (end user price) that fits entirely into a standard USB socket, with 1/8" of antenna sticking out. IOW, there is a way of solving it, whether it pushes the price point out of reason then becomes the issue since connector systems are the most expensive components on the boards these days.

[conrail98]

Now that I think about it more, I remember Duncan McRee talking about this when he was attempting the battery/WiFi no wiring product he had developed and demo'd on Crag Bisgeir's layout on a Model RailCast show. I'd have to go back and find that one to remember exactly what the issues he was finding with mass producing it but part of it was regulations and the exact conditions you were describing,

Phil

[DKS]

Besides... a DeLorean is more anachronistic than DCC... 

[LKOrailroad]

There's nothing I know of that would prevent anyone from adopting any sort of existing (or even developing new) technology. The stopping points are the cost to research/bundle/market the systems, and the high probability of a low uptake (not enough early adopter modelers) to amortize that cost. You can blue sky all kinds of fantastic train control systems. But how will they be funded? Figuring this out is far more important than product design.

And therein lies the real reason. It is easy to daydream all sorts of more modern technology to control model trains but to craft those dreams into a genuinely viable business case is a much higher mountain to climb. Proof is in the taste of the pudding they say. For all those with "great" ideas I still don't see anyone willing to risk a small fortune to see if they are the next big thing. 

[railnerd]

Besides... a DeLorean is more anachronistic than DCC... 

[John]

Me thinks my friend Zox likes to stir the pot a little    --- and over engineer something that doesn't need to be over engineered