G Scale Model Train Forum banner

Next Gen Interfaces - emerging schemas

148K views 183 replies 22 participants last post by  David293 
#1 ·
Over the holiday we had some fun evaluating interface schemas for the iPod Touch and Androids as submitted by a couple of my students for a colleague who requested them for introduction into his lineup for the New Year. Here are a couple that I found very simple and intuitive:






Hope to have many more designs coming through..

Cheers
Victor
 
See less See more
2
#2 ·
I still cannot understand the fascination with touch screen controls. If the controls are on the screen, then you have to LOOK at the screen to use them. There is NO tactile feedback. That's ok for getting a train running on a loop and then sitting back to watch, but useless for switching. I want to look at the trains, not the control.
 
#3 ·
Was just going to say, you need speed controls that you can use w/o looking at, and also you need feedback when you have bumped the speed up and down.

I have used both my smart phone and tablet... for running trains that only need occasional use, like running in an unobstructed loop on the flat, they are fine, but for any kinds of real "driving" of a train, it's worthless, and just a curiosity.

For programming a loco on DCC, it's convenient, and more portable than a laptop, there's a good application.

If you want to improve your interface, you need finer control of the speed at slow speeds, the circle does not work well, have tried it on other speed controls.

The multiple train lookup is ok, but you don't need a lot of screen space to display 5 locos... even the best "engineer" I know can barely run 2 trains at once in anything other than a "toy" layout.

I'd at least have the horn and bell buttons on all screens where you have a throttle.

Greg
 
#4 ·
The latest version of the schema has the following tabs:
Control - Trains - Tracks - Turnout - Lights - Sounds

The Horn/Bell on the Control tab would be more useful than on separate Light/Sound tabs - I agree.


Only two trains at once - in which world ? Try 16 over remote Command & Control.


Cheers
Victor

 
#5 ·
The post was about the interface on a smart phone.

So, the person using the smart phone would have to be very adept at running 2 trains at the same time.

If you were going to have the smart phone be the "nexus" for a complete automated command and control system, the amount of real estate is quite small for a programming interface.

If you are just going to start and stop a completely automated program, well all you need is an on off switch... that's not the type of train running I want to do.

Anyway, the WiThrottle and EngineDriver apps work ok for the people who use them, so their interfaces can be a starting point.

If you have a wide "throttle" bar on one side, you could probably change the train speed without having to look at the throttle, have it work as an incremental "knob" as opposed to using an absolute position on the screen for an absolute speed.

Regards, Greg
 
#6 ·


The Feedback layer has now been added to the Control tab. Speed feedback can driven from the rotary encoder if fitted for accuracy, otherwise from the motor driver feedback circuit. By having the Sound and Light controls on their own tabs, various sounds for different model locos can be selected although the debate of including at least one horn/bell combo on the main tab is still going on here.

We are approaching a design that is different from the Wiithrottle, EngineDriver or TouchCab as they all involve a broad DCC infrastructure to support them...Herein, the layout is inherently 'shareable' and the guest simply arrives with their own communicators (a.k.a smartphones). They download the local layout plan at the host facility wirelessly, select an available loco and begin navigating their trains. Also the design does not involve DCC, command stations or decoders - being a 100% wifi/radio alternative. Ideally suited for local clubs with large visitor attendance like our campus layouts where guests stay for an hour or two.

Cheers
Victor
 
#7 ·
Posted By VictorSpear on 28 Dec 2011 08:30 AM
Also the design does not involve DCC, command stations or decoders - being a 100% wifi/radio alternative. Ideally suited for local clubs with large visitor attendance like our campus layouts where guests stay for an hour or two.

Cheers
Victor



What hardware (inside the trains) are they using to the motors, lights, sounds, smoke, etc? Are they also providing motor speed feedback (or another method of detecting speed), or other useful features back to the UI? As you know I'm always looking for information on what hardware people are using inside the trains to do the controlling as this is still a grey area in my personal project.
 
#8 ·
More than 30 different MCU implementations are being tested by 4 University teams in different countries. The most popular family of MCUs in our field layouts is the ATMEL 32 bit AVR UC3 and last month the UC3L was brought into a design challenge competition theme proposal. The focus is picopower - around 165 microamps/Mhz in active mode ! This is usually combined with 'sleepwalking' features of the OBC with parrotpower for sound/light control. The target run time is 7.5+hrs/charge on certain models at 14.4 V NiMh. As mentioned in my earlier posts, location/speed/distance/velocity/acceleration/braking feedback is achieved with a fitted quadrature encoder (plus some integral calculus) where the UC3s have the encoder inputs built-in. There is an active University program too.


More recently, about a dozen TI ARM (Sitara) Cortex have arrived as 'Beaglebones'. Let's see what this gives us.


Cheers,
Victor
 
#9 ·
Strange, since one of the problems we do NOT have is power, either from the rails, or batteries.

The power used by the cpu is nothing compared to running motors, lights and smoke.

This does not sound like a project focusing on model trains.

The "university program" is a little misleading... Atmel, who maufactures chips, makes certain hardware available at a discount for use by university students.

That's like saying Apple or Microsoft has a university program.

I've designed several products using Atmel hardware, but none of what I have seen except your pictures of a screen mockup have anything to do with trains.

Do you have any links/reading on train oriented development?

What you need is low cost, wireless built in, high current outputs, and some form of networking.

Greg
 
#10 ·
I don't think you have kept up with the breadth and offering of the Atmel University program.Visit Cornell if you get a chance.



If your objective is to pull 50+ cars by supplying 24V/30Amps to the wired rails - sure you can use DCC plus 'boosters and 18 guage wiring' and blow the horns loudly all day :). We've moved far beyond the power sucking push model methods. In our world the modeling precedes the 'prototypes'. In an all battery/wireless radio environment you may note the 2 to 4 hrs of charging frequency in most of the current implementations, for example. Sure power is not a problem, if you want to load up trailing cars with fat batteries and run trains all day.


BTW that is not a mock screen at all.


Cheers,
Victor
 
#11 ·
You miss the point, I understand, I read the page... it's as I said... if it's different, point to the information. I was most likely designing devices with CPU's before you were born. I've done cutting edge stuff. There is no University class in a new train control system.

Cut through the bluster, technospeak, and provide some real information.

There's no question that Atmel helps people understand their hardware so the will use it when they get jobs... this is not a new or foreign concept.
Get back to the application of trains.

What is the blue blazes are you talking about here? "We've moved far beyond the power sucking push model methods."

Huh? you have trains that move without power? You have new electric motors that defy the laws of physics and can do work without consuming power?

You are going further out in left field here...

What galaxy is "your world" in? You state: "In our world the modeling precedes the 'prototypes'."

That statement is patently absurd... the idea of modeling is to mimic the prototype.... how can you precede a prototype from the 1940's in 2012?

Really, you need to stick to facts and what you can do to make a new control system.

That's not a mock screen? Well, send me the hardware and software, send it Fedex to my house, get it here in a week and I'll send you $1,000....

Yeah, right, the word "mockup" is not the same as "mock".... mock means fake in a way, mockup means like a prototype of the "look" in this case, but NOTHING behind it...

The $1,000 is waiting...


Make something, come with with a hardware design, analyze the costs, come up with a software architecture.

I've designed and implemented products before.


Greg
 
#12 ·
Posted By VictorSpear on 02 Jan 2012 08:34 AM
More than 30 different MCU implementations are being tested by 4 University teams in different countries. The most popular family of MCUs in our field layouts is the ATMEL 32 bit AVR UC3 and last month the UC3L was brought into a design challenge competition theme proposal. The focus is picopower - around 165 microamps/Mhz in active mode ! This is usually combined with 'sleepwalking' features of the OBC with parrotpower for sound/light control. The target run time is 7.5+hrs/charge on certain models at 14.4 V NiMh. As mentioned in my earlier posts, location/speed/distance/velocity/acceleration/braking feedback is achieved with a fitted quadrature encoder (plus some integral calculus) where the UC3s have the encoder inputs built-in. There is an active University program too.


More recently, about a dozen TI ARM (Sitara) Cortex have arrived as 'Beaglebones'. Let's see what this gives us.


Cheers,
Victor

Do you have any links/info on parrotpower? I tried searching a little but didn't see anything that would be replated to sound or light control from a MCU (don't know if parrotpower is either software or hardware).

Is part of the goal to create a hardware device that can be bought by modelers in either kit or assembled form or is the hardware and software mainly for university research and thus not going to be an actual product?


Also is the scale HO are are the various universities running G scale? I think MIT that you've referenced is HO only? I can see why battery running is more limiting on that scale.

Edit: Curious too, how is communication being achieved? I didn't think the Atmels had wifi? Are you sending binary, xml, or another format over sockets or a zigbee/zwave connection? (You mentioned wifi but I'm curious what hardware is handling this as I don't imagine you're runnning an embedded OS on the ATMEL are you?
 
#13 ·
Brandon,

I'm still framing a response to Greg's barrage ... but yours is simpler to answer :) ....you do have an open mind and do not dismiss what you don't know or have not (yet) heard about.


You may find the parrot event-driven DMA model interesting....if you haven't missed the point about Power Modeling Efficiency that I am making. Others will dismiss it as 'BScale' of course since their decoders will need over 3 amps to blow a horn at the rate they are going. Where is the modeling there, one may ask, if you want to stuff more amps into a wired train to make things happen ?


Much of the research is sponsored by manufacturers of course...our Universities just cannot afford it otherwise. Outdoor, solar/battery powered robotics are a top subject here. Our famous auto industry that went near-bankrupt twice is one shining example of the 'power is plenty, fool, don't worry about it' approach.


Take a good, hard, non-cynical look at this. Your next question should be " What sort of micro-amplifier can drive the output with stereo sound " ?



Cheers,
Victor
 
#14 ·
Greg,

"I was most likely designing devices with CPU's before you were born" -- Very likely to be the case..with all due respect....but age seems to cloud thy thinking now - methinks. It should be the other way around.


"I've done cutting edge stuff" - Why have you stopped ?

" There is no University class in a new train control system" --- sigh, which University have you visited recently ? I have a list of over 30...look at this list...just look at it..?


http://www.railml.org/web/index.php/supporters.html ( See if you can spot a University here)
http://www.city.ac.uk/engineering-maths/undergraduate/student-projects/model-train-set ( This is in an undergrad course )

http://www.tsd.org/papers/IEEE 1473-L Communications Protocol.pdf (Where do you think the next generation of this is being researched ?)
....

http://instruct1.cit.cornell.edu/courses/ee476/FinalProjects/ ( If you don't see model train applications here...you never will) Maybe Santa will.


"You have new electric motors that defy the laws of physics and can do work without consuming power" ( Consuming much less power and doing more. If you use the cheap $20 or less junk motors in the model trains that you now use....this may not be of interest).


"You are going further out in left field here.." And why not ? Almost everything you use today came from way out in left field. Not from mainstream status quo, I assure you.


What galaxy is "your world" in? You state: "In our world the modeling precedes the 'prototypes'." Yes, a paradigm shift. See the definition of 'prototype' in any way you choose. Then simply chose the way you want to interpret it. Model first or prototype first ? Here is a good statement that may help one out: China is the only country in the world to have a fully functioning, production-ready, German designed Maglev daily run. What came first, the model or the prototype ? Yes the implementation can be patently absurd to most.


Really, you need to stick to facts and what you can do to make a new control system. - Myself and a team of engineers from a known University are re-defining in a forthcoming IEEE paper, the very meaning of Command & Control. I'm uncertain if many can handle it. Hint: Follow the US Coast Guard model.


Mockup ? Here's something I call a mockup to people :) http://www.youtube.com/watch?v=9ZOqhPbj7rs


"The $1,000 is waiting..." No thanks, we are are talking about thousands of hours of student/faculty work here but your offer is generous. Look for a release at the end of March/April 2012 when the manuals and documentation are in place. When things are dismissed as BScale, it can wait. It will be a lot cheaper than that. Instead, send the money to MIT/Reas/Fry who support the Processing language. Incidentally, Fedex shouldn't be taking a week to get anywhere (my jab).


"I'm not intentionally giving you a hard time, but this is going from techno-bluster to BS." - No you are certainly not giving me a hard time, Sir...I have 200 minds that do more than that every day. But as for techno-bluster to Bscale ...do you really want me to respond ?


"Make something, come with with a hardware design, analyze the costs, come up with a software architecture." - It's what we do for a living but it may not fit some minds...


" I've designed and implemented products before."- Why have you stopped ? Just curious.


Cheers,
Victor
 
#16 ·
Since I am subscribed to this forum and really don't want to miss out on most of the discussions here, and I don't know how to unsubscribe from just this thread, I guess I will just have to endure it


At first, I found this to be of some interest. I'm a technical guy, but this stuff is way over my head, and I am assuming the same for 98% of MLS members. That is not to say it isn't still of interest to some, even though they don't understand it all. For me personally, the posts are way too long, way too complicated to be of any real interest, as far as real garden railroading goes,


I will continues to read the first two lines of each post with my delete key at the ready.
 
#17 ·
Sorry, didn't notice the edit on your post as I was toweling dry in my corner while refuting some jabs from another savant. Okay, here it is:

Communications: REST web services using URIs (not URLs, no SOAP), UDP, Zigbee. Core comm and network layer hardware (amongst several candidates) is here... - an SOC.


Cheers,
Victor
 
#18 ·
Del,

If I spent all my time responding to your jabs, yes I wouldn't have a job. Try CTRL+ALT+DEL ... it might be faster on your machine :)

Funny, you don't post much...you simply respond as "Nope", "BS..." etc...nice manufacturer vested style. I thought the forum was about sharing.


Cheers,

Vic.
 
#19 ·
A brief list of institutions that offer teaching, research, projects and industry JVs with scale models and real world applications

Australia
http://itee.uq.edu.au/~mint/
http://www.eng.monash.edu.au/railway/
http://www.qut.edu.au/study/courses/master-of-engineering-railway-infrastructure
http://itee.uq.edu.au/~aupec/aupec99/zou99.pdf

India
http://www.iitk.ac.in/me/Presentation/Mechanical.pdf

China (Beijing Jiaotong University, Beijing, China)
http://www.at0086.com/SJTU/Courseview.aspx?pid=78839
http://www.chinatefl.com/gansu/teach/lzru.html (tangshan Railway Institute)
http://en.cnki.com.cn/Article_en/CJFDTOTAL-TDXB201105017.htm
http://en.cnki.com.cn/Article_en/CJFDTOTAL-ZGTK200501022.htm
http://www.cnngo.com/shanghai/life/china-maglev-train-hits-1200-kph-636051

Germany
http://www.integratedsoft.com/CaseStudies/IFW-Dresden
http://www.uni-stuttgart.de/iev/

Korea
Hanyang Univ - Dept. of Transportation
http://www.hanyang.ac.kr/english/

Russia
Petersburg Government University of Railways
http://www.pgups.ru/

Swiss
http://www.ethz.ch/index_EN
http://www.ifor.math.ethz.ch/publications/2007_caimi_trainscheduling.pdf

UK
http://www.york.ac.uk/inst/irs/

US Universities

Several have railroad clubs (MIT, RIT..) where you can do your thesis/project

And also..
Carnegie Mellon/U of Michigan
www.ri.cmu.edu/


US (Next Gen)
Appalachian State University -
http://hydrail.org/
http://hydrail.org/node/44

...
...
NYU ITP with the Processing course (the core language for our models)

http://itp.nyu.edu/itp/
 
#20 ·
I'm finding this all very interesting, educational, and quite entertaining as a bonus. I'm a firm believer in using models to create prototypes and unless you take a stab at it you'll never make incremental progress...poking theoretical holes is really a waste of time.

Keith
 
#21 ·
Gentlemen,

QUIT THE PERSONAL ATTACKS OR THIS THREAD IS LOCKED!!!

Got it?

(This thread is too convoluted to try to weed out all the attacks from the good information to try to delete all the offending posts. Consider all parties involved equally warned.)



Del, you can uncheck the "subscribe to this thread" box at the top of the page to stop getting notices in your inbox.


K
 
#23 ·
Victor, will RailML be implemented on both sides or will it be a new communication protocol? Are there any white papers on how communication will work? Both data protocol as well as basic information such as if the consumer device talks directly to the loco or do the consumer device and the loco talk to a third party server that's managing the railway?
 
#24 ·
Ok, now that Squelch mode is turned on, conversations can be more focused on the subject of the thread - schemas. Yes, 2-way messaging from the start using the RailML open messaging format. If it works for the billion dollar industry consortium in the EU, it works for us. The sample RailML DTD examples provide wonderful ways to establish RESTful messaging between:

1. Object to Object ...{Loco to Loco, Loco to Switchyard, Loco to Station(s)}

2. Object to Coordinator (Handheld)
3. Object to Monitor (Repository of all known artifacts in the layout)


Our first model (Model 100A) provides for the simplest of loco-to-coordinator 2-way messaging.


Best.
Victor
 
#25 ·
I am confused.

Is this a new wireless control system designed for a model railroad application? If so, what scale is it intended for? Will it require users to buy new motors, gears, etc. in order to install in their current locomotives? Is it be battery or rail powered?

Please, no tecnho-jargon, no arguments. This is a model railroad forum, so I'd really like answers to real-world questions that a real-world model railroader can use to decide whether he is interested in this system. What does it do, what does it cost, and why is it better than what is already on the market?
 
#26 ·
Kenneth,

Points Noted.
The latest post was a direct response to a direct question from an active member involving communication modes. I was reminded by the requester to elaborate at least twice :)

To answer your specific question:

  1. The original post (Next Gen Interface Schemas) was to discuss taking a look at the emerging trend towards using commonly-available-household-devices for model railroading of any scale. Naturally, Large Scale is where it is easier to implement
  2. In an earlier post, Cougar I believe, had already mentioned Marklin's introduction of the phone/tablet interface at Nuremberg in early 2011. Why would the industry leader in our sport introduce this -> They are intensely studying the demand patterns of the age groups buying their products. (I know this. My colleagues worked on the current Marklin web site in German.)
  3. Touchscreens have entered the combat battlefield in the deserts today out of necessity and battlefield commanders insisting they be provisioned. See the General Dynamics implementation. Why ? Surely, there is a place for it in household garden railroading too. (See especially, the sunlight rendering issue addressal)
IF 1,2,3 above are not on the confusing siding, then the post is suggesting starting with a simple 4 button, scalable, shareable, inheritable interface that any 10 year old can also use (including mine).

The question then is about costs. To a Club registering a new member, providing the latest layouts should be a matter of emailing the layout file to them or downloading it at the club sites like we do - for free. It also contains the turnout settings, available tracks.
The Touch interface has various look/feel options that can be changed on the fly. Some like wheels, some sliders...some RED big buttons that vibrate when pushed.
And so on...we just provide a dynamic adaptation since it is possible now.

A handheld controller costs at least $150+ today, does it not ? While it does give you a 2 line display your guest may not have even that. Yet they may have one of these communicators (iphone, android). First quarter 2011, Gartner reported 428 million phones sold in the first quarter with smartphones accounting for 26%. That's a little over 111 million phones in one quarter, last year. Purists may argue, so what anyway ?

Anyway, the post was intended to argue what could be the cheapest, simplest, stupidest, 4 button controller that one could pass on to a friend when they visited your layout. Hope this clarifies ?

Cheers,
Victor
 
This is an older thread, you may not receive a response, and could be reviving an old thread. Please consider creating a new thread.
Top