If your manufacturing problem is primarily resource management and avoiding bottlenecks, then go for finite capacity scheduling, writes Geoff Lock.
“We envisage three key paybacks in the next 18 months,” says Arthur Densley, materials manager for the Plenty Group of his new finite capacity scheduling system. “We’ll have real time plant monitoring so we know where everything is in the plant; real time job progress so we know the status of every project; and accurate information will be available both internally and for the customer.”
This is an interesting and not madly expensive story. Plenty Group is an established manufacturer with 300 employees and a turnover of £25 million. It is made up of five operating centres which supply the process industries such as oil and gas, water, sewage and food processing. Four operating centres are product orientated, supplying mixers, pumps and filters; the fifth is a manufacturing centre producing parts and assemblies for the others.
Manufacturing has to cater for a wide range of products ranging from one inch filters to 20 metre long agitators weighing in excess of 20 tonnes. With a product range of hundreds, batch sizes from one upwards, lead times from weeks to more than a year and 30—40 projects being worked on at any time, Plenty Manufacturing has significant scheduling needs. The company now uses a scheduling system from Preactor, implemented to integrate seamlessly with existing working practices, while also making a significant improvement to manufacturing operation and giving customers the level of information they desire.
Densley looks back at how it used to be. “We were always having to try and serve five masters including our subcontract customers. At weekly production meetings a view was taken on the priority of each project based on the required delivery date, but this was very subjective and tended to be based on how close the project was to delivery and not on the availability of resources. In addition we were constantly getting requests from customers for dynamic schedules.” Not good. “Clearly we needed some sort of master scheduling system.”
Potential bottlenecks had already been identified as some of the CNC machines, and it was an unpalatable fact that projects of longer than 40 weeks duration tended to be slow at first, but require many more resources and effort nearer the due date. Understanding the roots of its problems led Plenty to conclude that it needed a project/resource scheduler that would handle the entire plant, also allowing it to monitor work in progress within the plant in real time. Furthermore it had to be integrated with the existing shop document system which had been developed in-house over an eight year period.
Having identified the key problems, Plenty set out to specify the specific core components that its solution would need to comprise. The overriding objective was to get more science into work scheduling without upsetting the lower level documentation systems that had been developed over time and fully accepted into the operation. This in essence meant being able to generate a master schedule built up from individual project schedules identified in the weekly production meetings. But it would also be flexible, and give customers the visibility they wanted.
“It was a thoroughly considered decision,” continues Densley. “More and more customers wanted to see the schedules and with our wide range of products, some off the shelf, some with delivery in one to two weeks, some requiring projects lasting many months, we didn’t have the means to show them this information. We also wanted to be able to pick up slippage and create priorities for activities.”
With the new system, Plenty has now moved from a situation where schedules were only produced if required, and typically only updated when the client next needed to see them, to one of controlled project planning and managed resources. Activities for each work area are analysed to generating what could be termed a ‘Bill of Resources’, similar to a Bill of Materials. The combination of these allows a reverse schedule to be generated for each job in which resources are identified and scheduled back from completion date. Information is also shared between the new scheduling system and machining and purchasing systems which makes fulfilment of works orders more efficient.
“Cultural change is usually by far the most difficult problem to manage,” says Mike Novels, managing director of Preactor. “It might take only two weeks to put some software in, but it takes much longer than that to have it working effectively. Where people have always done something in a particular way for a long time it will be hard to make changes. Just take for example the conflicts which can occur between piece-work and manufacturing to a schedule.”
Plenty has avoided problems of this nature as far as possible since there have been no significant changes to working practices. The Preactor system has been implemented to sit above established systems so people on the shop floor still use the shop document system they are used to. In this way nothing has been lost from previous investments and managing change has not been a problem.
Preactor’s system is almost fully implemented for the first product range, and will now move on to cover four or five more. It is also delivering the benefits it was installed for, by giving departments and customers better visibility of project plans. The drawing office now prioritises work based on dates generated by Preactor, and procurement and resourcing is based on the real needs of the project.
Golden Rules of Thumb
l Keep track of where you are and don’t get distracted. If you start with unclear objectives, it’s easy to get sidetracked. Leadership is needed when setting objectives and in implementation. Arthur Densley at Plenty took a leading role throughout: continuity is critically important.
l Work together on projects. Get all the departments involved around a table.
l If you’ve got your own IT department, involve it in the search for potential packages. It might be easier to buy a package and link it to existing systems rather than start again from scratch
l Start out with specific objectives. Plenty’s objective was to have a master schedule: but the objectives should be driven by manufacturing needs, not IT needs.
l The user must be the final arbiter. It’s all too easy to inflict systems on people who’ve been doing things in a particular way for a long time: they’re not necessarily wrong. Treat pressure from IT, accountants and others with an open mind, but also with caution: they don’t have to manufacture things.
l Don’t assume that the end users will understand a new system easily. Remember they’re the ones who will have to use it. Plenty took account of this neatly by retaining its old systems so the working environment didn’t change.
Author: Geoff Lock