A few months back I was asked to visit one of the big – really big – manufacturing software providers to talk to them about lean. They were particularly interested in lean accounting, and thought there might be a market out there they were missing. I might as well have made my presentation in Portuguese for all the good it did. Turned out to be a colossal waste of my time and their's. The notion that manufacturing management is experiencing a paradigm shift from the heady days in 1961 when Ollie Wight and IBM's Joe Orlicky were racing each other to install the first MRP systems is a notion too daunting for their conservative minds to grasp.
The big software companies are still convinced that all good manufacturing begins with driving a forecast to a master schedule, exploding and time phasing bills of material, then balancing lot sizes with set-up times to optimize inventory, and tracking the whole thing with standard costs. That is the baseline, according to their thinking, and every manufacturer should start with this approach to management. Pull systems and lean accounting are viewed as add-ons – something to hang on the side of the ERP system – but certainly not the core of a completely different approach to management.
So they all go to market with their MRP centered systems, Then look to their third party application developers to cobble together the lean things to hang on the side of their ERP systems. I told these folks they were sending their third party developers out with at least one hand tied behind their backs – that forcing the market to begin with MRP was severely limiting. I told them that MRP logic has a place in some lean enterprises, but not all and it is certainly not the core of manufacturing management in the 21st century. I told them there was a great deal of arrogance underlying their view that they knew more about manufacturing management than the manufacturing community, and that continuing to shove MRP down manufacturing's throat was a formula for eventual failure. I pointed out to them that, in most manufacturing companies, the most adept and prolific Excel users were the folks in supply chain and accounting – the two core users of MRP – which ought to tell them something about how well their logic was working. It is rare to see a want ad for a cost accountant that does not require advanced Excel skills – because cost accounting cannot get the information it needs from MRP. I might just as well have had the conversation with the cocker spaniel next door.
Now comes IBM with their "Smarter Planet" strategy – so blown up that the Prince of Wales is hosting a global summitto discuss the breathtaking implications of the concept. What is the 'Smarter Planet'? "The aim is to use software to analyse and predict trends in consumption, allowing organisations to save money and natural resources by adjusting supply to meet demand in real time." In plain English – the smarter planet is forecasting demand so it can drive through a master schedule … the same 'smartness' Joe Orlicky conjured up in Rock Island, Illinois 49 years ago. After years of pushing the old MRP monkey by dressing him in a silk suit, they are now trying to pawn off the 49 year old monkey by seating him on the royal throne.
Drill down into IBM's Smarter Planet and you will find that on the smart planet companies will be able to "drive economies of scale" and "implement volume price agreements". That sure sounds like the same old not-too-smart planet on which manufacturing foundered for years to me.
MRP logic certainly has a place in manufacturing. It is another tool in the tool kit, but it has been roundly disproven as the core of an integrated management system
Some day one of the big manufacturing solutions providers – SAP, Oracle, Microsoft, IBM – is going to have the courage to take the big step of building a new manufacturing platform from a clean sheet of paper … or maybe it will take a more aggressive start-up to redefine the industry. But the fact is that, for all the big talk about innovation and the marvels of our high tech culture, the software companies are lagging far behind manufacturing in advancing the state of the art, lacking vision and completely incabale of doing what they are supposed to do best – innovate.
david foster says
The proliferation of Excel-based applications (which are really usually a combination of Excel-based, email-based, and paper-based)goes beyond manufacturing companies and exists in businesses of all types. It’s not happening because people have a deep psychological need to become Excel experts, but rather because of the failure of both internal IT organizations and external software vendors to provide applications that really do what needs to be done.
Matt Wrye says
Amen! My company is currently (and by currently, I mean last month) implementing a major MRP system to replace all of the legacy systems from the past 40+ years. It has consolidated a lot of information which has been nice, but they are hanging their hat on the better demand forecasting module. “If we just had better forecasts………” I was seen as negative at times for trying to explain this isn’t what we want. I even cited many lean companies nationally, locally, and jobs I had in the past that they benchmarked from a lean perspective as getting away from MRP no matter what the system. Still, several million dollars in there is no turning back now. It WILL WORK! It is a shame because it is hindering some work that I am trying to get done.
David Dobrin says
I couldn’t agree more, but I also think the lean world needs to be clearer about why MRP doesn’t work and lean does. Admittedly, it’s an uphill battle, because most people who implement MRP declare victory and really believe their own declaration.
I saw this once again, a couple of months ago, when I did a post on what I call “brittle systems,” using MRP as example number one and asserting (if not arguing) that many (most?) MRP implementations don’t work. (Check out blog.b2banalysts.com.) The reactions I got were fascinating. “What do you meeeeen, MRP doesn’t work. Our MRP system works jest fine, thank you.”
How do I know they don’t work? Well, I’ve seen a lot of these systems, and in many of them, all you need to do is check the settings to see they’re not working. How do I convince others that they’re not working. I dunno.
david foster says
David Dobrin, I followed the link to your site and was interested in the post about the Cambridge public library system: “This guy simply couldn’t believe that the computer system that he had been given could have a flaw”
Several years ago, there was an example of the same sort of thinking–on the Washington Metrorail system–which had fatal consequences. Train operators were forbidden to switch from computer to manual control without approval of a controller, and this was generally denied, often with the comment “let the train do what it’s supposed to do.” This approach continued even though it was apparently known throughout the system that there were conditions under which the speed commands to the trains were being lost and when this happened, the local electronics defaulted to the *highest* permissible speed for a track segment..irrespective of weather conditions. On the day of the accident, the tracks were icy and slippery, the driver was denied permission to switch to manual, and the train slid into the train parked at the station, killing the driver.
I suspect that this kind of superstitious reverence for automated systems is depressingly common.
a reader says
I’ve found these comments very interesting. I always thought accountants were generally very happy with their MRP software whereas most manufacturing users such as myself have had to cobble excel workarounds to get the data that we really need.
Phil Coy says
MRP worked when the business model was make to stock and companies could afford to maintain high and stable inventory levels. MRP as the replenishment engine to produce inventory was right for the time. Times have changed.
I’m not holding my breath waiting for the big ERP companies to get a fresh vision of manufacturing and end their reliance on the old MRP planning model. To them, lean is just one checkbox on the checklist of the hundreds of features and functions that they support. So long as their customers and prospects from the manufacturing community allow them to provide only a little kanban or some lean accounting, they won’t spend another dime of their R&D.
What about waiting for a start-up with sufficient vision? Attracting investment money for a software player with an entirely new manufacturing approach is very difficult. The hot areas are “green”, healthcare, and all things gaming, mobile, and web. The new ERP players into cloud computing like NetSuite are not looking to reinvent the MRP standard – merely reset the price point using the cloud. Venture money is risk-averse – what’s the risk of going up against the prevailing and accepted practice of MRP endorsed by professional organizations like APICS and implemented by the deep-pocketed giants SAP, Oracle, and Microsoft.
Jim Fernandez says
Bill:
I’m not clear here. Did you go to speak with the software company to get them to adopt lean principles in their business? Or did you go to specifically discuss how MRP and lean don’t work together? Did the discussion start with how to use lean in software developement? And then it evolved into a discussion on MRP systems?
Another point. I don’t think true software developers are at all inovative. I don’t think they ever have a “vision”. The best software developers observe what is happening, figure out what is needed and then work with the customer to develope the software to do the job.
If we want an MRP system that works with lean, I think we need some empty headed software developers to go into a great lean running company and see if they there are any software solutions that would help the company with their lean journey.
*********************************************
Jim,
The purpose was to explain the basic inability of traditional MRP/ERP software to support lean accounting and lean operating (factory scheduling) principles.
Bill