Theory of Constraints and Scheduling Software

A Valid Schedule: The Holy Grail of Manufacturing?

To give this some context: it’s now 2008, I first scheduled a plant (a food processing plant) in 1973, and I was part of the revolution that saw scheduling software emerge from the domain of million-dollar mainframe computers in the 1970’s to become affordable to smaller manufacturers on first mini-computers, then PC’s, and now with today’s various configurations of PC’s, servers, and networks.

That’s 35 years.

In the 70’s, a scheduling pioneer called Oliver Wight called a valid production schedule “The Holy Grail of manufacturing.” He proposed some scheduling logic for use in factories.

In the 80’s, another scheduling pioneer called Eli Goldratt said of schedules in general, including those generated by the Oliver Wight approach “Cut out the middleman with your schedules – connect your computer printers directly to the shredders and save everyone a lot of time.”

Both were right; and today, my partner and I still make a very good living from our consulting because 99% of all the development in the world of scheduling since then has failed to provide 99% of manufacturers with a valid schedule.

Rod and I do provide our clients with a valid schedule, and the benefits are always massive … and so we earn our income.

Scheduling Truths

There are just a few things important to understand about scheduling, and scheduling software – the two go hand-in-hand today, of course.

The scheduling logic in most software today isn’t far removed from the logic in the 1970’s. All that’s happened is that the massive leap in computing power has enabled computers to generate and regenerate schedules very quickly, and to show extremely impressive screen displays while doing so.
The problem with this is, that the scheduling logic embedded in 99% of these software packages cannot generate valid schedules. Period. All the forward scheduling, back scheduling, optimization etc makes very little difference … the people who designed the software do not understand what makes schedules invalid, and therefore how to “immunize” schedules against the reality. So the schedules they generate simply don’t work. This is true no matter whether the scheduling software is part of an ERP package or stand-alone, from the most reputable ERP suppliers in the world or a local company, whether it’s labeled “finite scheduling” or “infinite scheduling,” or whether it’s labeled something else … with one exception.
So, what we now have is invalid schedules being generated and regenerated at the speed of light.

I know of one software vendor, for example, with literally hundreds (perhaps thousands) of users in North America … and virtually every customer bought their system based on its perceived scheduling abilities.

I know sales people who sold many of those systems. I know an installer who helped customers implement dozens of systems. And both will admit that fewer than 5% of the purchasers actually ever use the software for scheduling, once implemented.

Why not? Well, the reasons they’ll give vary, but they usually include (in this sequence, often over a few months):

Inaccurate or incomplete data.
Lack of management commitment to the project.
A workforce that lacks the discipline to follow schedules, and supervisors who lack the back-bone to enforce the schedules.

All of these may genuinely exist, of course. Regardless, what typically happens is that the company eventually gives up, and simply uses the system (in addition to using it for their financials) to maintain bills of materials, “explode” some bills, record inventory transactions and balances, and track purchase and work orders through their cycle.

As I say, all these excuses may exist … but they don’t stand in the way of a valid schedule! People who don’t understand scheduling, simply think they do.

The only consistent solution we’ve found – Theory of Constraints

The only scheduling approach we have encountered that consistently generates valid schedules is the Drum-Buffer-Rope logic of the Theory of Constraints. When you understand it, you’ll have an “Aha!” experience when it comes to scheduling a plant.

And for almost 20 years, my partner and I have implemented DBR logic without needing DBR-dedicated 3rd party software. We’ve used whiteboards (i.e. basically a manual system) or – most commonly – Excel spreadsheets.

When a company has an MRP or ERP or home-grown system in place, we’ve taken advantage of it – for us, it’s just a source of the data we need to construct valid schedules using Excel.

And don’t walk away with the impression that Excel = “very limited.” We scheduled a plant with 2700 skus, 600 intermediate products and 200 raw materials, using a spreadsheet. Most of our clients today are still scheduling using the spreadsheets we put into place years ago.

But this year … we’re favoring some DBR software. It simply provides too much functionality for us to ignore, it provides far more flexibility, it’s a more robust solution and best of all – it speeds-up our implementations so our clients get results, faster, AND it ends up costing clients LESS than if we were to custom-develop a Spreadsheet solution.

If you’d like to discuss this topic, take the next step – schedule a call.