Software simulation of laser job

[Shane Turner]

I get many inquiries to do custom jobs. It sure would save a lot of time if I could run a simulation of the job ( to get a job time and hence price ) instead of having to wait till the laser is free or guestimating for a quote.
Does anyone know if there is a way to do this?

-Shane

[Gary Hair]

I have been searching for almost a year for the same thing. I found someone who was working on a program for that but after I sent him a boatload of time data for my machine I haven't heard anything back.

I think that if you determine the time it takes to laser various sized areas at various dpi settings you should be able to guesstimate pretty closely. You will always be off a bit, hopefully you guestimate high - you can always tell the customer you over-estimated and it's less than you thought - what a hero you will be to them!

Gary

[Shane Turner]

True :)
I usually guestimate pretty well. But oftentimes when I'm out working on the laser, Karen will be manning ( personing?) the fort and need to give quotes. Would be great for her to be able to run a file through a programe and get a time.
Looks like i might have to write one myself. If somehow I could get feedback from the drivers I should be able to get an exact time. Diving in the deep end....

[Jim Watkins]

I think if you run your Epilog with the top door open, it will run the job without firing the laser which would give you the timing without having to burn your stock. I may be wrong, but when I was getting my demonstration, the man lifted the lid to show that the laser would stop firing, but the job kept going. When he shut it, it picked up where the lid closed showing a gap in the job where the door had been opened.

Maybe the others who know more can confirm or correct me.

[Stephen Beckham]

Jim - True, door open and the job runs without firing the laser. Used for tracing with the red dot, getting a total time on project or to act as an emergency stop firing. The head will travel the full speed and distance.

I started a chart where I could try some 9th grade algebra to figure out a "man-chart" to help with. I got a lot of data put aside and then opened the store. Never had time to go back to it. They tell you to get everything set up before you start selling in a storefront, because you'll never have time after your first customer...

[Shane Turner]

That's how I currently get my times for jobs. though I was wishing for someway of NOT having to run the job to get an accurate time. For a complicated job that can take 10-15 minutes to get a time, and that's time i could be using on doing production work.
I'm trying to think of some of way capturing the drive data sent to the laser. That way I could just PRINT the job and get an instant run time on it. Anyone know anything about print drivers and such?

[Lee DeRaud]

Warning: much handwaving follows.

I think if you can get the basic "X milliseconds to raster one sweep at speed Y" equation (more likely, lookup table) nailed down, it should be fairly easy to have a Corel VBA macro that computes the pseudo-area (sum of the net sweep lengths) of the envelope of the rastered shape. (That assumes you know what shape your particular machine's driver actually sweeps out for non-convex shapes. Mine is both non-obvious and counter-intuitive.:eek: )

[Shane Turner]

Hehe indeed. Somehow that made sense to me. :eek:
Which is why I was hoping to capture the output from the print driver, as I assume all that information is already programmed into it ( or maybe it's in the firmware ) Shouldn't have to reinvent the wheel me thinks.... hopes...

[Richard Rumancik]

I don't understand why the printer drivers for the laser systems don't have a built-in module for estimating job time, especially for raster engraving jobs. The PhotoGrav program has had a time-estimate module built-in and it was released many years ago. It would not be convenient (or cheap) to use this program exclusively for job estimation, but tells me it should be doable. (I do not use it regularly but I did some tests a few years ago and found it was pretty reasonable in its estimates.)


I think if you can get the basic "X milliseconds to raster one sweep at speed Y" equation (more likely, lookup table) nailed down, it should be fairly easy to have a Corel VBA macro that computes the pseudo-area (sum of the net sweep lengths) of the envelope of the rastered shape.

I suspect Lee is on the right track. I think that the "tight" left-right envelope of the raster area is the constant velocity zone, so if you could add up all the "lengths" of these constant velocity passes and know the speed (in ips) then you can calculate the elapsed time. (time = distance travelled /speed). You have to omit completely "white" lines as the laser will skip them. But to get total time you need to add the turn-around time on each pass. The "turn-around time" is the time it takes the laser head to over-travel, (where it decelerates, stops, and accelerates for the return pass)

This will probably have to be found experimentally. The PhotoGrav manual tells you how to determine it. It will also tell you how you can figure out the actual speed of the laser (in ips) by using a test plot. I generally use only 100% speed but if using more than one speed you might need a table as Lee suggests.

If you read some of the background info in the PhotoGrav manual you might get some insight into how it could be done. I wouldn't be able to write a macro but if someone was to attempt this, take a look at PhotoGrav first.

(I have been unable to verify that the raster zone is a constant-velocity zone. When I read the PhotoGrav manual it suggests to me that the rastering occurs in the constant speed zone. But in another thread it was suggested to me by Peck that the laser accelerates and raster plots at the same time. I doubt that mine is. I suppose this is theoretically possible but certainly not trivial to implement. Perhaps the newer lasers can do this trick. If this kind of algorithm is used on a laser, then it will be difficult to write your own macro.)

I don't like using the laser to do raster time estimates, as it puts a lot of wear and tear on the motors and bearings, as well as tying up the machine in "real-time." But I still do it when needed as it would not be practicable to use PhotoGrav to do estimates. It only accepts .bmp files as well, which is a limitation, and it would not be easy to test different dpi scenerios for job time.

[Bob Cole]

I know the new Universal machines (PLS series) have a built-in "job" estimator built into the driver. The ULS rep did a demo of it and I was surprised how long it took to "estimate". Maybe this was due to a slow computer but would have thought the estimator would be pretty close to instant. It does the estimating in software and doesn't require the machine to be running (no dry run needed).

I purchased the X series so left back in the stone ages for driver capability.

I would think this would be a nice add-on for the manufactures either as a purchase add-on or included in the driver.

[Ed Lang]

My ShopBot CNC machine has a good estimator on the cut time. It also has a preview screen for the cutting. I can see how long and exactly where the cut(s) will be. Also I can turn on the tool simulator and see the tool moving around. This take a lot more time as it runs, simulates in real time.

I don't know how to write something like this, but I sure could use it!

[Mike Ireland]

The new Universal systems (VLS, PLS and ILS) all have a job time estimator built into the driver. The time it takes to estimate depends on the processor speed in the PC.

[Larry Bratton]

I was reading the manual that came with Photograv and I came across some information on how to calculate the engraving time. It's too lengthy to post here but the information in on Page A3.1 (Appendix 3) entitled "Calculation Procedure for "Turn Time". They give a simple method and one that is more complex. Maybe this will be useful.

[Harry Radaza]

I have been looking for exact same thing. just input the design (bitmap or any picture format) and out comes the run time. However, to no avail.

The closest I found is this www.fastimation.com (http://www.fastimation.com) but reading from its site and a quick phone call to the maker suggests that you input your estimated engraving area. Not very accurate I suppose. It has to do more with costing since it also allows you to enter your per minute charges.

So back to the same problem. Luckily, I hired a graphic artist with extensive background in programming (I don't) not too long ago. I can have him write the program with your help. After completion of the program, I would be willing to offer this to creek members at a discounted price and even offer it for free to those contributors in this thread who helped give advice in the making of this program.

The only factor would be the time, especially with the xmas season coming up. So he will probably be in stop and go mode in writing this.

To start this, what data would I need ? Machine speeds as discussed above ? But this would only be good for my machines ( Mercury and Versalaser ). Maybe we can get the data needed from the different manufacturers and it would simply be a pull down menu for users? Then from there, just load the image onto the software and out comes the run time.

Then maybe in the future, add our individual per minute charges together with markup %'s and it would be a complete package.

Any other ideas to help me get started with this ? I am thinking maybe by next week my programmer can start...

[Gary Hair]

To start this, what data would I need ?

I wouldn't rely on the info from the manufacturers, they are going to skew it to their advantage. There was a person doing this almost a year ago and asked me to run some speed tests on my laser and send the results to him - he kinda dissapeared though. What he had me do was run varying line widths and lengths at various speeds to determine the run time. This data was supposed to give him the info he needed to calculate the true speed that the laser worked, not what the mfg said it would do.
Logically, it is a pretty easy thing to do, programatically it's not so easy. I have done quite a bit of programming in the past but this is beyond my capabilities. If you want the info I gathered, and the files I used, email me and I would be happy to send them to you - if I can find them that is...

Gary

[Zvi Grinberg]

I know the new Universal machines (PLS series) have a built-in "job" estimator built into the driver. The ULS rep did a demo of it and I was surprised how long it took to "estimate". Maybe this was due to a slow computer but would have thought the estimator would be pretty close to instant. It does the estimating in software and doesn't require the machine to be running (no dry run needed).

The PLS driver estimates the job, by simulating the process. If the job includes raster, then it would take long time. Vector jobs would be estimated pretty fast.



I purchased the X series so left back in the stone ages for driver capability.

I would think this would be a nice add-on for the manufactures either as a purchase add-on or included in the driver.

Well' this is not possible because th PLS series is designed with a different architecture (hardware, firmware, software, control etc). Therefore the new driver cannot beimplemented in the old X or other series.

[Bob Davis]

While a quick calculator would certainly be handy for pricing one-off large jobs, I usually find it's much more important to know how much the market will bear. I detest the idea of putting an hourly or by-the-minute rate on stuff we do, and I'm likely to spend more time on the setup and design than what the lasers take. If they're rastering I'll be doing something else anyway. (I learnt the hard way not to walk away when cutting acrylic...)

I'm always surprised when I read that somebody will use a shop rate like $60 an hour to price a job, or charge $1 a minute of actual laser time as the basis for pricing a job. The right price is the maximum amount the customer will pay. They are paying for a result, they don't want and can't use our time, and really that sort of calculation is just a simplistic cover-all for a way to set prices. Naturally, the type of customer you have and service you supply will be a big factor, but for most of our uses the laser time is almost irrelevant.

[Gary Hair]

While a quick calculator would certainly be handy for pricing one-off large jobs, I usually find it's much more important to know how much the market will bear. I detest the idea of putting an hourly or by-the-minute rate on stuff we do, and I'm likely to spend more time on the setup and design than what the lasers take. If they're rastering I'll be doing something else anyway. (I learnt the hard way not to walk away when cutting acrylic...)

I'm always surprised when I read that somebody will use a shop rate like $60 an hour to price a job, or charge $1 a minute of actual laser time as the basis for pricing a job. The right price is the maximum amount the customer will pay. They are paying for a result, they don't want and can't use our time, and really that sort of calculation is just a simplistic cover-all for a way to set prices. Naturally, the type of customer you have and service you supply will be a big factor, but for most of our uses the laser time is almost irrelevant.

Bob,
I agree in part with what you say. I do use my $120/hour shop rate to work on a minimum price - that may be very different than the selling price. If you don't know what you need to get, at a minimum, to be a profitable business then you will be in trouble pretty quickly. I'm not implying you don't know that, or that everyone is running a business, but if you are then you need to know this number.

I always "adjust" my price with a "perceived value" or "going rate" adjustment. That is how I sell a name badge that takes less than a minute, $2.00 at my rate, and has $0.65 in materials, and sell it for $18.00 + a setup charge of $15.00 - that's what the customer expects to pay, it is the "value" they place on the item.

Gary