The future of model railroading
Through a hole in the fabric of space/time, an article from the March 2012 Model Railroader has somehow appeared on Terry Thompson’s PC. We bring it to you now for a glimpse of the future of model railroading.
It’s 7pm November 4th, 2011, and the regular crew are arriving for operating night.
Many are carrying the same Digitrax DT400 they have been taking to layout visits for years. The South Pacific Lines uses only simple knob-and-switch UT5 and Walthers FRED3 throttles, or PDAs, or the new JMRI train simulator interface. All programming is done on computer with JMRI DecoderPro and all switches are thrown by CTC. The throttles just run trains. But I can’t get the guys to give up their favourite devices that have programming capabilities too. I just make them promise not to program anything. Part of the trick to keeping a complex system like the South Pacific Lines stable is strict change control, just like the business IT people do.
Something else they carry that we didn’t see a few years ago are PDAs, phones and laptops equipped with JMRI and either Trainz or TrainSim. More of that later.
Also walking up the drive are a couple of visitors I invited along for a look. In the past we wouldn’t normally welcome visitors as the train room isn’t that big and the aisles couldn’t cope with railfans. Nowadays that is not such as issue. In fact the layout room can be eerily empty at times, as we will see.
The crew spread out around the house. The TV room, Jack’s playroom and the dining table all get converted into crew space for the evening. They log in to the wireless network with their phones or computers, and sign on to the seniority system for the evening’s tasks. Those with radio throttles test the signal. Some are still using wired throttles so they find a seat near a Loconet jack, which I have wired throughout the house.
Peter gets the dispatcher’s desk. In front of him are a prototypical paper timetable graph fresh off the printer, and a not-so-prototypical flat panel display with a representation of a CTC panel on the screen (so much cheaper than a physical CTC panel).
We start the LocoNet fast clock. On the train server, the emulator reads the timetable and starts the SF010 early morning freight drag. It shows up on the CTC panel in the departure track at San Francisco. Tim has signed on as engineer so he logs on to Trainz and selects the cab of SF010 and “climbs aboard”. The dispatcher gives him clearance, the signal on Tim’s PC screen turns green and he eases the train out. The lounge rumbles quietly as the subwoofer adds some real vibration to the loco sounds filling Tim’s headphones, and David’s headphones as he rides alongside on his own PC.
In the layout room SF010 sits unmoved on a staging track, as San Francisco is not a modelled part of the layout (who has a train room that big?). But in Tim’s virtual world every track of the division yard is there. The mainline is there too, but just as we do on the layout, selective compression has been applied to shorten it down to the interesting bits so Tim can get to Oakland within fast clock time and without getting bored.
As he approaches Oakland, we come into the modelled sector of the track. Tim likes technology: he prefers to run the train from the PC. His screen switches from Trainz simulation to the forward in-cab video camera. In the train room, the Tortoises whirr automatically to clear a path and the Athearn SD40s of SF010 come to life and run from staging onto the modelled layout. David ducks into the train room to act as Tim’s crew: setting manual points, watching for derails and doing the coupling.
David passes me as I’m talking to the visitors – the only others in the train room at the moment. The subwoofers in the train room thump with non-directional bass that really fills out the diesel bellow coming from speakers on the SD40s, as SP010 bursts from between the buildings onto the elevated tracks over the port area.
Peter on CTC clicks the mouse to set the points and signals for the Oakland loop, and Tim uses his on-screen throttle to bring the model train to a halt. This new JMRI/Sim interface that both Trainz and TrainSim have adopted means that he doesn’t have to switch to a Digitrax throttle to run the real HO train – he can keep controlling it just as he did the virtual version.
It is time for the Oakland switcher to get to work for the day. Wayne has never taken to the new technology. He prefers traditional DCC to get close to the model trains. So he puts his drink down and walks into the train room, old DT400 in hand, and plugs in to bring the switcher out. The availability of half-inch micro-electrostatic speakers means I’ve recently been able to install a sound system in the little 0-6-0 with the speaker tucked under the cab roof. It would still sound a bit “tinny” if it were not for the room’s sound system that backs it up with some bass. Wayne cuts a block of cars off the back of SP010.
David lets Tim know there are a couple of cars to pick up before they leave Oakland. Tim switches to the rear-view camera and backs the train up watching David’s hand signals in the train room. There are purists who say the view doesn’t look quite right with a huge 1:1 scale hand looming over the tracks giving the “back up” signal but I think the realism is great. Here we have an engineer who can’t see anything but the view out his rear window, relying on signals from his brakeman to back on to couplers of some cars that are all but hidden beyond the end of the train.
Tim calls up Peter for clearance and rolls out onto the main again. As the model train disappears off-layout into staging, Tim switches back to Trainz to run the virtual train on to Richmond and beyond.
Wayne is busy making up train SP023, the Alameda turn. I excuse myself from the visitors for a moment to play hostler in Oakland. My son Jack is doing the job tonight because most of the work is over early and he can go to bed. He is having a problem with loco 127, the 0-8-0 assigned to the Alameda turn. I grab a UT5R wireless throttle and run 127 onto the repair track beside the roundhouse.
I flick the switch on the workbench to hook the repair track up to the workbench’s own Zephyr and little Toshiba laptop. (I like to have a portable – or at least luggable – test bench). This laptop holds the loco roster and is the only thing allowed to be used for programming decoders. I know that JMRI can share the roster on a network, but I’ve got change control bolted down real tight and I’m not comfortable about letting programming out of the bag just yet until we think it through. Once I’ve sorted 127 out I release the repair track back to layout control for Jack to run the loco onto the ready track for Wayne.
SF02, the Daylight, is due out of San Francisco at 8am. The schedule emulator starts the train running on the server. No-one in our crew is too interested in running the passenger train over the simulated part of the layout, so it runs quietly by itself with a virtual engineer on the PC. Peter watches it moving along the CTC panel, and he sets the signals and points for it to give it a clear run. There are a couple of locals and through freights that never actually appear on the modelled layout portion but they run on the emulator and keep Peter busy holding them off the main for the varnish to come through on time.
As SF02 nears Oakland, Tony is always keen to drive the big Alcos, so he selects the consist on his DT400R and settles back to watch the in-cab view on his phone. I warn the visitors to pick a good spot to railfan the crack train as it roars across the layout. Tony gives them a blast on the horn as he sees them flash past on the screen, giant figures standing beyond the edge of the world. They nearly jump out of their skins. Neither of them has been to a layout with 100 watts of surround sound under the benchwork.
The fun bit over with and the models back in staging, Tony leaves the train to the emulator to run it north to Richmond. Only Peter watches it now on the CTC panel.
Meanwhile our switching nut, Keith, has signed on for the Alameda turn, train 23. He uses a PDA running a JMRI throttle of his own design (using the JMRI throttle development kit and a good friend who speaks Java). He selects loco 131 via the PDA’s Bluetooth wireless and goes to the train room. Everyone is different and Keith still doesn’t like in-cab camera, which is lucky because I haven’t installed it in 131 yet. Something to do with the mortgage, according to my long-suffering wife.
Speaking of investment in essential technologies, I invite the two visitors to tear themselves away from the train room and come see our latest project. The only free space I could find was down in the basement where my first layout used to be. Here we have built a half-size replica of the engineer’s side of a diesel cab interior. A replica diesel throttle is linked to LocoNet. Now that 40-inch flat panel plasma displays are so cheap, I have picked up three from a remainder bin and installed them as forward, side and rear windows. One of the new three-camera video units is on its way from Japan. When it arrives it will go in the Daylight lead Alco. Then maybe I can get someone interested in running the Daylight over the whole line and not just the modelled bit. There is enough room in the basement to do a steam loco cab too, but it will be a few years before we attempt modelling that in 1:2 scale. Besides I am still struggling to hide the camera heads in the cab frame on an HO steam loco. I envy the larger scale modellers where the cameras are now all but undetectable from a few feet away. Any day now I guess we will see factory-installed triple-head cameras.
A call comes over the headset that brings us hot-foot back upstairs. The emulator’s randomiser has decided to derail a unit train just north of Richmond. Peter is coping well on the dispatcher’s desk but things are choking up in the staging. We need to make up a scenario to clear a few trains onto the train elevator and back down to the storage tracks below the benchwork. David grabs a PC and starts acting as Richmond tower to help Peter re-route traffic while a few of us confer in the train room, where it is quiet right now.
We conclude we should redirect a couple of freights east through San Jose instead, so they will never appear on the modelled layout. Peter agrees so while he makes the changes in the emulator and CTC, we load the modelled trains onto the elevator and drop them down into storage. Now the staging frees up and the modelled trains can keep up with the emulation.
Some might argue that the modelled layout restricts the simulation, especially because of the artificial bottleneck caused by the staging tracks. We think it wouldn’t be the same if the whole layout was virtual. Half the fun is knowing a train is approaching, and railfanning it as it appears in our location, just like the real thing. It is even more powerful for all those modellers switching to O and G scale now that they model only one town, though we are staying with HO for now. Simulation has freed us up to model a single location in a fairly small space while still operating an entire division of the Southern Pacific. It doesn’t get better than that.