[jbaz]

I have only evaluated the low to mid tier priced CAM packages to date, and frankly I'm not sure if there are any CAM packages that have evolved to this point yet, but I would be surprised if none of the higher tier packages had this type of automation yet, at least to a certain degree.

What I find most time consuming is testing different cutting and tool path strategies and the combinations of them to find the most optimal strategy. This is even more evident when it is a part requiring 3 or more tools to make the part because it has features that can only be cut by certain sized cutters.

What I am looking for, because it would be a game changer for me personally because I do a lot of one-off, would be where you input your tool library like most CAM packages so it knows what tooling you have available. Then, you load in your model and the size of your stock and it will create an automated strategy that would be most ideal based on the quality of the finish/accuracy, which can be tuned.

To be able to load up a part, and the software automatically selects the best tools to cut the part based on the tools I have available, using fairly optimal roughing and finishing for the part... wow, that would speed things up a ton for me at least. Perhaps the issues I have with the cheaper solutions aren't as evident in the higher end packages, but I always find myself having to define specific regions for cutters because it has medium and small slots and pockets and it just can add up to a big strategy list and a lot of time. It is a breeze for simpler parts that just need a rough and finish pass using 1 or 2 tools for the entire part, but the parts that need multiple end mills for different features on the part is when it would be nice to have the PC do the work for me.

So I guess the question is are there are CAM packages that have these "Smart" or automated features that can make a pretty good tool path just by knowing what tools are in the library, the model and stock/material? Which, if any, seem to be the most advanced in this regard?

[hoisee]

I use CAMWorks so can only introduce some features of this software. You can define your strategy in its Access Database. For example, you can define a pocket strategy that will do a series of operations whose parameters and tools and cutting depths can be assigned in DB. As for the tool selection, you can assign a diameter range, or a constant diameter, or a specific tool ID then the software will select the tools based on your criteria. If you organize your TechDB well, it can achieve your requirement.

[jbaz]

That sounds pretty interesting, I will look more into that for sure.

It sounds like you understood what I was talking about, and it sounds like CAMWorks is obviously a step closer to what I was talking about compared the packages I have been using. But after reading my original post I thought I could do a better job of clarifying what I was after in for others to weigh in on.

What I am trying to see is if any the higher end CAMs out there are doing even more of the work, where it is actually analyzing what tools you have available in the library and picking the most optimal strategy to cut the part from the tooling you have available based on the part you are trying achieve. So if I load a model in the CAM and pick the stock size (and possibly tweaking the speed vs quality parameters), the software automatically builds a fairly optimal cutting strategy from start to finish.

It would obviously be great when, at least for not super complex parts, to have software evolved enough that is akin to opening a document on your PC and clicking the print button and the software can get comparable results to an experienced individual in terms of quality of the resulting part. I have no doubt software can optimize the fastest strategy by testing multiple ones against each other when you tell it what tool and area to machine, but the heavier logic would be in knowing what tool to select for different features in the part and what strategy would create the desired finish/quality balanced against time.

Anyone else know of packages that are doing anything comparable or making steps towards that?

[jbaz]

I guess my Google-Fu wasn't very good. It looks like in Mastercam they refer to this as Feature Based Machining. Would those of you who are fortunate enough to have used the higher end offerings say that Mastercam is the leader in this type of automatic toolpath creation, or are there other ahead of the game? Are there are any mid-tier offerings out there that offer a similar feature, albeit not nearly as robust or useful on complex parts? I don't normally deal in complex parts, mostly just 3 axis single side machining, but they often require 4 or more different end mills to efficiently complete it. Not a big deal in and of itself, but since I do a lot of one-off and variations it gets monotonous setting it up each time in CAM when it would be nice for the CAM to do more of the lifting, who doesn't like lifting less. =)

[hoisee]

The strategy in CAMWorks is determined by user, not the software. So this is not what you want. As for mastercam, I had serveyed it several months ago. It looks like a software that is about 20 years ago....... If you do a lot of one-off including designing then in my opinion you should not go for it, but you can call a demo
We do almost everything one-off from design to manufacturing, so integration and file management is important. If your business including designing the part, you should use an integrated cadcam solution.

[jbaz]

Thanks for the info. It seems like FeatureCAM is somewhere geared towards what I was talking about, but it is more towards "features" as in holes, pockets... typical 2.5d stuff. We have some 2.5d stuff and some 3d stuff (mostly single sided 3 axis machining, some 2 sided "flip" work, but has 3d contours). I think maybe the automation is less important but I just need to step up to a slightly more intuitive package. After looking at some of the videos from other offerings I can see how I could create the tool paths with less work using them.

[TheCamMan]

The problem with what you’re looking for is your term “Optimized”. The concept of optimized falls back on the difference between how computers work and how the mind works. Especially when it comes to 3D milling. When you’re looking at 3D mill there is a lot of conceptual thinking that goes into what is optimal and what isn’t. Just adding a little bump in the surface (or face) could totally change the cutting strategy you would use.
When talking about 2D there are some that do an ok job at guessing how you want to cut things, but even there you end up having to add in the conceptual thought process yourself. Consider a pocket with a round island in it. You can set up how you would mill it and it would work most of the time. But what happens when you hit a pocket that’s too close to the wall for your rougher to fit? You end up having to check everything the system has done very carefully in order to determine if the toolpath is acceptical.
In MHO you won’t find a system for quite a while that will be able to overcome the conceptual thought issue.

[zool]

TheCamMan stated: In MHO you won’t find a system for quite a while that will be able to overcome the conceptual thought issue.


I use CAMBAM, which is not a real intuitive cam package, though very versatile and deals well with complex shapes; it has a steep learning curve, but then relatively easy to use thanks to a great forum and constant incremental ongoing improvements by its creator and add-ons by users.

I am a woodworker and machine 3D parts, and all my jobs are one-offs.

TheCamMan has astutely stated the difficulty of your finding a cam package that can think as well as you. You think in a comprehensive, non-linear, and abstract fashion, i.e., you are very inventive [its real obvious!!!], while a cam package is linear and calculating. You know what you what, the cam package does what its been told.

While wood is much more malleable than metals, I have a box full of busted bits and a few pieces of wood with ugly cuts in them on the wall to remind me of what happens when I depended on a cam package too much.

Programing g-code to complete hogging out, waterline rough cuts, and waterline finish cuts, each with a different bit, is not simple, at least for me, and I shutter to think of depending on a software program to make decision about how I want a piece to look when it is done.

Granted my work tends to deal with appearances a lot, though all the pieces do have to fit together in the end, while your metal working no doubt deals with tolerances, or broadly: fit and finish.

Software driven machine tools can do a lot for makers, whether mass-producing or one-offs. Those that do mass-production spend hundreds-of-thousand on developing and verifying g-code BEFORE production starts. In the end though, like TheCamMan, I believe that those of us that produce one-offs are still going to have to complete significant reviews of tool paths and g-code interactions between tools and materials for a long time, if not indefinitely.

[jbaz]

Thanks for the replies. The more I look closer at the higher end packages the more it seems that they are trending towards more automation like I described, but it isn't entirely there yet. I know when it comes to holes/drilling/tapping and the like it is hard to get away from needing to define strategies. But when it comes to just creating a 3D object with external features/milling only, using strictly flat and ball mills it seems like some of them are very close to being able to automatically build an efficient strategy for a desired finish/tolerance with multiple ops by choosing the right cutters and tool path parameters automatically based on the machines material removal abilities. I can't say for sure because I haven't personally used them, but just watching some videos and reading articles it looks like it may not be such a fantasy at some point.

[victorofga]

jbaz

it boiling down to your knowledge.. with a ""low"" end cam you might able to producing what others creating with 20K package..

how much need to be automatized? it also pending how much job you have on hand.. if theres a lot then you can go any high end program...
if theres not a lot job coming then start with a low end program like cambam, and after you up n running then you can go for an expenses cam..

just think practical..