Out of our lovely discussion in the iGen3 area...

How do you estimate run time for your digital press for any given job?

Do you use rated speed or the manufacturer's guidelines?

Do you use the job log and build averages?

Do you factor things like quality or ink coverage into the mix?

Do all your jobs run at your estimated run time?

Do you factor in a spoilage number?

Do you have a minimium or maximum run time?


I'm curious to see what others do for calculating a job's run time, we use a combination of a percentage of rated speed and type of job it is to determine run time. Sadly, sometimes we can be way off and either get the job done really quickly (not the norm, believe me) or due to something out of our control, like jams, the job could run way over the estimated run time.

just as importantly as what you use for factors in run time is the order you prioritize them and the relative weight you assign each factor.

typically i have to look at:

rated speed run time.
paper.
total hours from first item to determine maintenance windows effected
quality requirement of job
finishing options
vi / static
etc.

estimating run time is one of the most difficult things to do in the print world. there are so many factors both in and out of your control that it can quickly get away from you.

I remember a really cool feature on the Creo DFE for the iGen3 that would estimate remaining time / elapsed time not sure if it is still around or not.

It would be a really nice feature for a DFE to have a way to tell you how long a job should run, taking into consideration all the factors above. Average the time it has been running and adjust accordingly. Would not help estimating, but you would know if you are going to be way off faster then finding out your are half way through the job when the MIS estimate time is up.

I wonder with the XML status info the Nexpress feeds out if that could be parsed to build this kind of information. I know the Creo was outputting some sort of status XML file, so wonder if that could be done for that side too. I don't know if DocuSP or the Indigo DFE can do this or has a status file that could be parsed.

Maybe a script that would poll the server every so often, check the status of the job, timestamp it, get a meter read and based on how many impressions needed, and if they provided the total pages / sets done (to factor out spoilage) you could extrapolate the remaining run time. Hmm, might be something worth looking at, then store that information back to the MIS package for costing and later estimating updates.

a key factor would be the amount of data sampled to get valid estimates. in a short time that estimate can be incredibly inaccurate. even if all things are running well if your sample does not include a cycle of loading and or unloading paper, it could be statistically insignificant.

I agree, but feed in the MIS estimate data and then you can get a ballpark, and then just a +/- of the data and statistically subsample the data to extract a more accurate runtime estiamte during the run.

For most runs, under a couple of hours, probably not that big of a deal, but for the longer, 10+ hour runs this would be a great status tool to see how far along you are on a job compared to estimate and provide a realtime estimate for completion based on current running conditions.

such a tool would be nice. it would also need to be able to gather data from more than one press for situations where you are running multiple hours on multiple devices.

Usually rated speed + 5 min per 1000 pages for jams and refills.