It depends on how you use it!

I've been using CAD to design layouts since 1998 here are some things I've learned...

• It's a LOT easier to design a layout than it is to build it. What seems easy on paper or on your computer monitor is likely to require much time (and possibly skill) to construct.
• I still find it difficult to envision exactly how the track plan in my computer will look when built. The actual trackage for the industries in Browning on the BC&SJ bears only a passing resemblence to the planned industry spur tracks.
• I almost always see better ways to lay out spur tracks and sometimes yard throats in full size on actual benchwork. This happened for sure in Browning.
• When I was designing the second BC&SJ I used huge amounts of time creating terrain, buildings, roads, bridges, etc. (and admiring them in 3D) for track plan versions that I later discarded in favor of a different plan. Time is the only non-renewable resource.
• It's cheap to add a turnout in a CAD program. It will set you back at least $5 if you scratch build it and more like $20 if you buy turnouts. And that doesn't include switch stands, switch machines, linkages, or control panels. It's easy to design something your Chief Financial Officer will frown upon. It's easy to design a layout way beyond one's financial means. Or beyond one's available time for the hobby.
• I thought that setting my CAD program's 3D viewpoint elevation to match my eyeball height would let me "walk around" in a virtual version of the multi-deck second BC&SJ track plan checking out lower deck visibility below and upper deck. In practice what seemed OK visibility and clearance-wise in CAD didn't work out so well once I was well into construction.
• If you're not already familiar with CAD you'll need to climb up a (potentially steep) learning curve.
• I can't recommend the freebie Atlas Right-Track cad program (unless you'll be building a layout entirely of snap-track components). You'll probably need to pay $$ to get a program that will be useful for you (I'm sure there are a number of folks who are using alternative CAD programs that are available for free - please chime in here).
• If you don't frequently save your work and the CAD program crashes, you're likely to become slightly (yeah, right!) upset. It may take a number of crashes and accompanying lost work before you learn to save more frequently. Maybe AutoCad is crash proof (if you have thousands of $$$ for your CAD software and the expertise to configure it for model railroad puposes). I don't know of a hobby program that can make that claim though crashes are generally few these days.
• Finally, and this is probably the most important detail, if you don't already know how to design a good track plan, a CAD program won't design it for you. It won't obviate the need to know what you are doing when it comes to designing useable yards, feasible yet interesting industrial areas, selecting LDEs. If you'd do a poor job of creating a track plan with pencil and paper, the same is likely true with CAD (excepting you won't end up with too sharp radii or turnouts!)

Wow! After all that it looks like CAD is a terrible idea. To that let me borrow a phrase the apostle Paul seemed fond of - "May it never be!"

• If you use CAD in an honest manner, it will keep you from designing yards with #2 ladders.
• It will keep you honest with track center to center distances.
• It will prevent sub-standard radii curves.
• It will compute exact elevations on grades and allow you to verify there is sufficient railhead to railhead clearance when one track cross above another.
• If you print it out 1" to the foot, you can use an architect's scale to take arbitrary dimensions from the plans.
• It will let you design benchwork and give you the lengths of each piece of wood used in the support structure.
• Even if you're like me and seldom built a layout exactly the way it was designed, CAD will let you know that the track concept you have in mind will fit the space allocated for it.
• A simple line-drawing of building foot prints can assure that there is enough room next to spur tracks for the desired industry (less important when designing and scratch building your own industries)
• Accurate and quick calculation of the length of your mainline, passing siding capacity, spur track capacity.
• Some CAD programs will give you a shopping list of what you need - how much track is required, how many turnouts of each type. Handy for estimating cost.
• With a bit of practice, I could doodle basic track plan investigations about as fast as I could sketch them with pencil and paper - and be confident I haven't designed in too tight radii or turnouts or too steep grades.
• In tight quarters when every 1/4" counts, CAD will let you determine whether a track configuration is buildable or not and if printed full size provides a construction template for that trackwork.
• I've gotten lots of mileage out of my 3rd Planit license. In addition to having designed myriad track plans I used it to annoy the architect when my house/train room were being designed, make full size templates for bridge and building construction (just print 'em and put down some double-side scotch tape then layout out the sticks directly on the plans), I've produced 3D illustrations for a number of my MRH articles with it - I've even produced a few animations for videos with it.
• It's easy to share highly accurate track plans with your friends over the internet.

But, should you get lost in CAD during your layout design process and spend the next 10 years designing a layout, adding tons of 3D features and wandering around in your virtual train room - if you're having a good time who's to say you've got it wrong - though Trainz might be a better choice for you than a layout design CAD program?

Charlie

CAD Like...

I used CorelDraw to create my trackplan. It has cad like accuracy and I know the program inside and out. It was natural for me to begin there.

I was using CorelDraw to design our future home at the time, and I was in no rush to start laying track since there was no house yet. I worked on my layout plan right inside the house plan - in fact, in many ways, the house was designed around the layout.

I never used paper or pencil to conceptualize my BNML, but I did go thru at least 5 distinct iterations of plans before settling on one - and that plan was also subject to many more revisions.

I did include roads and trees and even vehicles in my plans and I have to agree:

"It's easy to fall in love with a CAD creation"

I found myself endlessly tweaking my plan and just staring at it, dreaming of how it would look and operate. It's true, I was in LOVE with a drawing. wow. there's a tough revelation.

BUT with that said, I feel the same way about the layout that is unfolding in front of me and I am pleased that all the details that I labored over are turning out just as I had imagined.

...And with all the detail that exists in my current plan I am now stitching together my photos for the backdrop right inside CorelDraw, on top of my plan, and THIS IS FANTASTIC! - I am able to get the pictures sized and organized so they will fit without a bunch of print - cut - reprint episodes.

This picture shows individual photos beginning to be placed...

Do I regret spending so much time on my plan? no way. not even a bit.

I like CAD, but within reason

I think you've covered the pros and cons fairly well.  For me, CAD (I prefer XTrackCAD - http ://www.xtrkcad.org/Wikka/HomePage) allows me to quickly design any layout for any space, and never run out of paper.  More importantly, it allows me to easily change any and all aspects of the design without leaving a paper covered in corrections and erasures.  It also makes it easy to design multiple completely different layouts for the same space - I draw the room and any physical constraints, save it, and then make multiple versions, each saved with a different file name.

However, I am always aware of the danger of spending too much time making a plan too detailed.  I have often said that no layout plan survives actual construction, so there's no point in getting too detailed.  I am on the XTrackCAD mailing list, and whenever someone asks about modeling a specific structure or piece of rolling stock, I try to politely remind them that the plan is just a draft, a test to see what fits and what flows well in a given space.  Making sure that the box car you put on the plan has the right number of roof ribs is a complete waste of time.  If you want to explore virtual model railroading, there are better options than most layout planning programs.

Cad Use

I have to go along with all the comments put forth so far. And then I started to build my RR and like Charlie, what was on paper didn't last too long when it came to town locations. I am free lansing. If using a LDS cad is the way to go. I started using track componets and or turnout photocopies. Look at it for a few days or weeks. Tweek it, look some more and then lay track and use it and change it some more. And all that includes using structure templates too which might just be a box.

Jack

I've been looking 30 years

For a computer drawing program that would give me the freedom and creative flexibility I get from pencil and paper. (As of February anyway--I waited for the Mac 512 to come out.) I've tried everything from MacPaint to TurboCad and Illustrator, currently I work in EZDraw. I am at heart a doodler. I've learned to doodle on graph paper, so my drawings have enough discipline to be buildable.  To me, a CAD drawing is a type of final product, something for a book illustration, or for plans others will use.

To generate multiple versions of a layout, I use a quad ruled note book. It's easy to transfer fixed dimensions and layout elements from page to page and create variations on several themes. When I have something interesting, I scale up and do a dimensioned drawing--architect's rule and compass keep me honest. (I learned drafting at an early age.) My drawings aren't as pretty as a CAD drawing, but they are just as buildable.

I'm sure the specialized rr CAD programs produce fine results, but frankly, I'll be lucky to build four layouts in what remains of my life, and they will be separated by enough time that I would have to learn a new version for each. I'd rather make two or three mistakes in wood and foam than two every three minutes on a computer.

pqe

ps. Some of you may know that I am a composer. I have exactly the same attitude toward scoring programs like Finale and Sibelius. They make beautiful final copy, but I do the real work with pencil and paper.

Another use for 3rd Planit...

Although not the best tool for it, I just used 3rd Planit to design the Toh Jct tower display panel showing turnout positions, repeated signals, and occupancy. Somehow, although I haven't designed a layout in a while I keep finding uses for CAD.

Charlie

clarification

As I might be the person who's comments caused this thread, I think I should clarify how I use/used CAD in the layout design process.

I free hand sketched the C&W until I was happy with the track plan. This included side elevation and sections. When I was happy with the sketches - their prototypical fidelity, photographic opportunity and operational design - I put the track plan into CADRail to confirm the plan, order the required switches and keep myself honest. That was as far as I took the design.  I didn't get into 3D nor get detailed about scenery...or run trains on the plan!  Looking back at the initial sketches the track plan has changed maybe 10% max since then. 

However once the track plan *was* in CAD I found it invaluable for the following:

- designing benchwork and all its elements which are very complicated on the C&W.  CAD helped immeasurably here.  Things like swing and lift gates, helixes, indexing table, tilted open grid benchwork, support hangers for the upper levels, etc.

- locating Tortoise switch machines under the sub roadbed clear of benchwork and light fixtures

- cutting curved backdrops, their supports and their location

- scaling and designing bridges, piers and abutments and the benchwork to support these

- designing critical structures (e.g. the barge slips which, as on the prototype, actually ride on rails sloping down into the lakes, snowsheds, etc.).  

- realistically considering and developing those 10% changes to the track plan

So I don't believe CAD is necessary, nor perhaps even valuable, to design a layout unless it is quite complicated.  In fact my normal method for designing layouts for others, including multi deck layouts, is to put the room plan in CAD, drop in control points, and ruling curves and aisle ways as guidelines, and then print out multiple copies of the CAD file upon which I free hand sketch different track plans into the space.

md

I still doodle on paper...

...and get a general idea of what I might want to do.  I'm no great track planner so a lot of times I "borrow" concepts I like from published plans.  In the end though I do most of my actual design work with a handful of turnouts and some flex track. I only use SCARM,  which is my program of choice, to refine things and make a presentable looking plan to share with others. My way  of doing things is probably not the way to design a larger layout  so if I ever get the space to build my dream layout then I have three favorite published plans that are candidates for it and I'll probably hire a pro to design me something original based off of one them. Either that or to modify one of them for my needs.

Michael

 "While it can definitely

"While it can definitely help with making sure everything fits, if your goal is getting a layout designed and built CAD can turn into an enormous time sink."

Yeah, if one is not sure how to design a layout I think the time spent learning cad would be better spent studying track plans and figuring out the good and the bad of each one until they know what will work and what won't. A copy of John Armstrong's "track planning for realistic operation"  and a few issues of "Model railroad planning" annuals would go a long way in building up a knowledge base of layout design. Once the concept is understood then the need for cad or not will be apparent. Most layouts should not be so complicated to need cad and if they are perhaps the layout is too complicated?  A person who really needs cad should already know they need cad before they get to the point of asking about it :> ) ........DaveB 

I came across something

I came across something called cadrail. It supposed to help you design a layout and you can run trains on it too. I have never used it but it might be what your looking for.

...

This is like asking if plastics and plastic injection was a good thing back in the 1950s.

In the end, it's a not a choice or a matter of "good idea or bad idea."  It's where the technology is going.  If you make use of it or not is up to you.

Personally, after using computer software for drawing layouts for a good decade now, I would never do a final layout track plan any other way.  You CAN get exact precise layouts now, in the planning stages, the old "the track that gets laid never quite fits what's drawn on paper" is a myth nowadays.

Uses for CAD ...

... in model railroad design ...

1. Helps to ensure that what is designed will fit the space.  It is far too easy to "fudge" a curve radius or space required for a turnout/crossover.  A CAD program keeps everything in scale.

2. Helps to ensure clearances can be met; both horizontal and more importantly the verticals clearances if the design requires them.  It becomes easy to check the grades required to meet those clearances as well.

3. Running trains ... perhaps the #1 reason for using a CAD program designed for model railroaders.  You can actually "operate" your design before cutting a single piece of wood.  This will help identify potential problems BEFORE the track work is installed.

Which program?  Many good ones about.  I started with CADRAIL and then migrated to 3rd PlanIt before finally settling on XTrackCAD.  XTrackCAD is FREE and does everything that I NEED a CAD program to do.  (3rd PlanIt is an awesome 3-D railroad CAD program that even includes 3-D trains.  The only "problem" with it was, that I was spending ALL my time designing (and running) "virtual layouts"!)  

...

I take my XtrackCad plans to sketchup as image files, size them to scale, and then trace the trackplan using Sketchup's tools.  The final result is my plan in 3D, which is pretty cool for mocking up large complex arrangements.

Theoretically I could then print the design out on a 3D printer and have a physical scale model of my future layout, whatever it may be, in the same way the pros used to talk about building a scale model of the future layout.

Good for independent work, not very useful for collaboration

“It's easy to share highly accurate track plans with your friends over the internet.” 

I respectfully disagree. I would argue that it’s actually almost impossible to truly share plans created in layout CAD programs—at least, not in a way that's useful. Sure, some products let you export your creation to a static file format such as PDF or JPG, but that’s about as useful as printing out a hard copy and faxing it. Your friend across town or across the country can’t very well open your PDF or JPG file up on their computer, add a siding or adjust a curve, and shoot it back to you for discussion and further editing. 

Even the most cursory market research would show that we are an extraordinarily collaboration-oriented bunch who would be well-served by software interoperability and a standard layout data format (exhibit A: MRH’s Trackplan Database), yet the layout CAD software industry is, for the most part, offering products designed around the needs of mid-twentieth-century hermits. Based on recent conversations with some leading model railroaders whose perspectives I very much appreciate and respect, I understand that most of the CAD software vendors are very small enterprises, and that many of them would have to rebuild their products from scratch if they had to incorporate a data format standard, and that such a burden could even be enough to put some of them out of business. Part of me is highly sympathetic to those arguments. But another part of me is pretty annoyed that as of 2014, antiquated, proprietary thinking still dominates this software market niche and is preventing us all from collaborating on a richer, more robust, more value-adding level. 

So to return to the question of whether CAD is a good thing (I'm really not trying to hijack the thread, I promise!), IMHO the answer is yes but with the caveat that it's nowhere near as good as it could—make that should—be. I’ve tried several packages on for size and found a couple of them to be useful tools for exploring layout concepts, test-fitting potential alignments and arrangements of track and other layout elements, and generating alternate versions of a plan to see which is most likely to meet my design objectives. But the lack of interoperability limits my ability to use CAD software to design collaboratively, which prevents me from using CAD as far into the design process as I otherwise might. Which has some up-sides as well, particularly in that it drives us to go hands-on earlier rather than later and reminds us that at the end of the day it's all about the trains.

Designing a standard format

Want a standard file format?  I think it's a great goal, and here's what I think would have to be done to make it happen.

First, figure out everything which has to be recorded about a track plan.  At the very minimum, I would suggest that it have to contain XYZ coordinates for every bit of track, or coordinates for the ends of track segments and some kind of data (radius, at least) about the track in between those points.  Figure out how to include properties like track gauge and whether it is hidden or not.  Throw in some kind of line drawing (with width and color) definition, and you're most of the way there.  Make sure to leave some room for custom definitions which may or may not be used, and which may safely be ignored by other programs.

Now take that format you've just spent quite a bit of time developing, and make it public.  Share every aspect of it, and describe in detail how to use the format.  Chances are, you won't do all that unless you're writing some sort of layout design software, in which case you're going to have to give away your work to your competitors.  Even when you do, most of them will ignore you, as will almost everybody else.  If you keep at it, and if your software is good enough that a lot of people start using it, then you'll slowly see a shift.  New versions of old programs may include the ability to input or output your new format.  People may even write little programs to convert layout files from other formats into yours.

All it takes is a model railroader and computer programmer with the passion to see the project through.  that, and enough skill to make a really good piece of software that people want.  And a bit of luck.

What I find amusing is that it's a lot like the RailPro vs. DCC debate.  Do you stick with a standard that everyone understands, but which may restrict your ability to do fancy new things, or do you scrap the standard and head out on your own?  Both have advantages and disadvantages, and both options will frustrate someobody.

"But another part of me is

"But another part of me is pretty annoyed that as of 2014, antiquated, proprietary thinking still dominates this software market niche and is preventing us all from collaborating on a richer, more robust, more value-adding level."

  Drawing layouts  is no different than any other drawing task so maybe it would be easier to eliminate the niche factor and look into general use cad programs? The might already be a decent inexpensive universal program out there that could be used for layout design if interchange of layout data is desired?  I've never looked into the possibilities because I can design anything I want to build on a sheet of paper but if I did want to run a cad program I'd probably look for an old copy of autocad , I had autocad 13 on a previous computer and it worked fine...DaveB

Absolutly Good for Collaboration

Being able to share CAD layout designs is a good thing, and I do.

Most of the RR CAD programs mentioned here have the ability to export and import DXF, DWG or similar standards. Working with other modelers, I regularly move files between XTrCAD and 3rd PlanIt. I have not tried with SCARM (emf or Enhanced Meta File exports), but understand that it is possible. XTrkCAD track layouts can also be moved into JMRI PanelPro. Most versions of AutoCAD and SolidWorks support DXF or DWG formats.

That is not to say that there are no problems when exporting and importing files. As with most of the graphics standards (eg. IGES, ANSI, ISO, EN) the DXF & DWG formats do not support all of the functions and capabilities provided by these packages (the basic reasons that different data formats are used); therefore, moving files from one CAD platform to another is rarely a smooth process. As, and for very good reasons, none of the major CAD system software packages (eg. AutoCAD, SolidWorks) use the same data structures either, expecting small niche players to harmonize their offerings is a bit shortsighted.

Remember that Martin Fischer, a model railroader and computer programmer with the passion to see the project through, developed XTrkCAD. But also read the development history of XTrkCAD, any you will have a better understanding as to why CAD programs do not use the same data structures. 

If you enjoy using CAD, I

If you enjoy using CAD, I don't see it as being a detriment to your hobby interests. Afterall, we do this for fun. I use Solidworks 40-55 hours a week lately. I would like to design my next layout entirely on it, but at the end of the day I really don't feel like doing the CAD work any more. I have done a couple sample layout benchwork plans just to see how it looks, but adding trackwork is a whole different ball of wax that I have not tackled yet. 

I do find 3D modeling to be a great brainstorming tool, but there is a steep learning curve for a hobbyist that doesn't already have experience with the software.

Playing with a Mushroom-

Multi-deck with Helix-

Single Deck add-on to another layout- 

"I do find 3D modeling to be

"I do find 3D modeling to be a great brainstorming tool, but there is a steep learning curve for a hobbyist that doesn't already have experience with the software."

 Plus the other big problem is these things only draw the lines, they don't tell someone if it's good design or not so a newbie buying a program and thinking it will help him build a better layout quite mis-leading.  To really be of value to an inexperienced modeler the programs would need some basic knowledge built into them. Something like "here's my room, I want to model the Pennsy at Delmarva in 1950, what should I draw?" Lacking that any Cad program is just making prettier lines than a guy with a pencil ....DaveB

Pencil and CAD same same

Can you provide an example of software in any field that can do this? Software is a tool same as the pencil. Neither provide the solution, just a means of recording the solution.

NO! It Is Bad

I prefer the old school set of tools. They do not require power. They are cheap, and the only resources are new lead and erasers. You never have to pay for an update, or deal with system crashes. Charlie, Im old fashioned and require knobs and switches on my throttle too.

I keep them stored next to my bamboo Slide Rule.

Regards, John

The theoretical understanding of the world, which is the aim of philosophy, is not a matter of great practical importance to animals, or to savages, or even to most civilized men.<> Bertrand Russell

I love XTrkCAD

I don't know if I could design my current layout plan without XTrkCAD or something similar. If I did it on paper, it would take much, much longer and not be nearly as accurate. I've printed out portions of my plan 1:1 and laid track right over the drawings and it has been right on. My plan is fairly complex with three levels, two helices, a crossover, and some  up and over trackage that I'd hate to try to draw by hand. Maybe the biggest advantage is how easy it is to make changes. Save the current version, make your changes, save under a different name, decide which is better. I am modeling prototype tracks and have had to go back and make lots and lots of changes as I get better information and that would be really frustrating with pen and paper. It is so much faster, and easier, to make good accurate drawings with CAD.  

The other really cool thing about XTrkCAD is that you can run trains on your plan to see how it really operates. Not only is this a great tool during your design phase that you could never duplicate on paper, but it is also just plain fun. Now, that may slow down your construction process, but since this is all about fun anyway, who cares?  

And now for Artificial Intelligence

But Alan, when that AI system conceptualizes and designs the layout, and the robot elves build the system, it will be up and running in no time. However, just when you pick up the throttle to begin running the new layout, a Terminator zaps in, throws you in the trash dump and enjoys running your layout.

It's a wave!!!!!

Warn Dave, not me. He is the one that suggested it. 

I made it through the design phase without robots showing up. If I had used pencil I might still be at it and in fear of the graphene shavings from the pencil electrocuting me next time I plug in an appliance in my bare feet.