In February, we reviewed management-information-system (MIS) trends and selection guidelines (“Making the best MIS choice”), followed by some user case studies in April (“Real-wold MIS”). In this issue, we asked Donald H. Goldman, the principal of Consult-Ware, a Pembroke, MA-based consultancy specializing in MIS issues, to discuss some key computer-assisted scheduling (CAS) challenges. Prism-USA (Pembroke, MA) recently announced Goldman has joined the company as senior project manager.

　　Why don't more printers use CAS?

　　Many lack formal scheduling practices — going from no system to CAS is a quantum leap. The central person managing the schedule must be aware of all the jobs in the plant, including priorities, due dates and component/production status (e.g., paper, ink, prepress) in process. Also, there must be a workflow step where every job is planned (estimated with hours) and job-status information is automatically collected and flowed into the CAS module. Without these elements in place, getting information into the system can take just as long as it would with a manual system.

　　Some vendors claim to have a scheduling module, but it is no more than a loading program where the hours are posted to each cost center. Other systems are too complex or inflexible, making it inconvenient to reschedule or try “what-if” scenarios.

　　Why do printers often expedite rather than actually schedule jobs?

　　Commercial printers are being pressured by their customers to shorten production times, and consequently go into expedite mode, pushing jobs through the plant, relying on overtime, moving jobs around, doing partial runs and other tricks. In such cases, a “schedule” board or spreadsheets are used to track jobs and build lists that are used to prioritize work during daily production meetings.

　　But this isn't true scheduling — it's dynamic loading. While it lets the printer know what work awaits, it often ignores scheduling methods that can optimize equipment utilization, improve operational efficiencies and maintain on-time job deliveries.

　　How do you define loading, dynamic loading and true dynamic scheduling?

　　Loading is the forecasting of sold production hours leading to providing departmental backlogs along with showing job status and tracking. It's the minimum printers should do to manage their workloads. But loading isn't synonymous with efficiently organizing the work for production or optimizing resources.

　　Dynamic loading is a hybrid approach — the system lists the jobs by resource and due date, but it is up to the scheduler to move the jobs around or make other adjustments manually. Many printers prefer this method of scheduling, since it more closely approximates a schedule board.

　　True dynamic scheduling is the prioritizing and sequencing of the plant load/forecast. The scheduler starts by looking at the due date and then, working backward, determines the due date for each prerequisite operation. Scheduling identifies any conflicts that will delay on-time delivery and offers a dynamic way to shift jobs to other equipment to see the impact on the plant. Dynamic scheduling systems apply business rules that help automate the decision process. The smarter the CAS, the less the scheduler has to intervene or run “what if…” scenarios. A CAS module also graphically represents the schedule, often allowing the scheduler to drag and drop their scheduling adjustments and otherwise see the results of dynamic changes.

　　What specialty/type of plant is best suited for CAS?

　　Typically, larger plants with multiple days allowed to produce their work are the best candidates. They also can show a good ROI, since they can take the biggest advantage of optimizing their resources through the intelligent grouping of tasks, and will minimize scheduling department staffs by eliminating redundant, manual functions. Midsize to larger plants also estimate all of their work, so the hours feeding into the schedule are readily available.

　　Smaller plants, especially on-demand types, are probably best suited to a dynamic loading system where jobs are sequenced based on due dates. It should not take more time to schedule the job than to run it.

　　Do you see the Internet as a potential aid to the scheduling process?

　　In general, the Internet is a great tool to improve external and internal communications with sales and production. It is good for communicating with customers via e-mail and for supporting production activities like proofing. But, it is less clear that giving customers access to the schedule is desirable. A few years ago, Xerox surveyed commercial and in-plant printers on the desirability of providing access to job-status information to the printer's customers. The commercial printers said, “No way!” while the in-plant printers thought it was a good idea. It depends on the way you want to build the relationship between yourself and your customers along with the culture of your organization.

　　Some printers have stumbled when attempting to implement electronic scheduling because their standards were geared toward selling rather than manufacturing — estimated hours aren't necessarily accurate and production standards don't reflect actual productivity. Any solutions?

　　A printer should only have one set of standards. The standards should reflect reality and the general practices of the company. In estimating, printers should be doing cost estimates and sell the job based on market pricing. Tinkering with run speeds or makeready times to artificially affect pricing prevents the smooth flow of the work from estimate to production planning to scheduling. The scheduler will still be doing some fine-tuning, since the estimate often includes washups or additional times for perfecting-press changeovers. A good scheduler will find ways to sequence the work for optimal use of resources.

　　Also, some printers erroneously think they can't schedule non-machine operations such as prepress. Time allowances (standards) can be computed — as they often are on the estimate — so there is some idea of the amount of work to be done. When you define work in hours, it can be scheduled.

　　What are the advantages of CAS vs. a manual board?

　　Labor reduction is the biggest advantage. In manual, card-based scheduling, cards must be prepared, edited and organized. Next, the cards are placed on a wall board and must be adjusted at least once a day. Also, the scheduler needs to walk the plant periodically to determine the status of the scheduled work. In a plant handling 20 or more new jobs per day, it can take several people to keep up with the workload. Most board-based scheduling systems also require printed documentation — often spreadsheets — for distribution and discussion with customer service, sales, production and departmental supervisors.

　　In a CAS module, all of the necessary data is available to the scheduler electronically with production plans (estimates) passing on hours, quantities, material requirements, purchasing history and other pertinent information. Job-status and -production information are captured automatically from the shop floor through the data-collection module.

　　Are there any drawbacks to using both an electronic system as well as a manual board?

　　It is redundant and costly to maintain both. If the CAS is a true scheduling module with good graphics, there is no need to have a manual board, too. Some companies probably ran a parallel board while implementing their CASs and never stopped.

　　It's been said that that the initial electronic scheduling implementation is easy — the true challenge is monitoring and adjusting it as it evolves. Do you agree?

　　Properly implemented and staffed CAS does not present any challenge to monitor or maintain. Scheduling requires good start-to-finish workflow discipline and a scheduling manager who is empowered to make decisions based on current job information from sales, customer service, the shop floor and others.

　　Too many companies think scheduling is a clerical job and assign junior employees to print the schedule, find the information and make entries — but don't empower them to make decisions.

　　Do you have any tips for conducting an effective production meeting?

　　If I had my way there would not be any production meetings — or at least not the kind I have often seen. It is amazing how many things are discussed at these meetings that could be resolved with one-to-one communication as an event occurs or through an inquiry to the plant's MIS. The worst meetings are those where the production manager goes over every job in the shop, including those that have been delivered. Questions on the status of proofs or materials or delivery instructions are usually the key topics. These meetings seem to go on forever.

　　We need to learn how to input and retrieve job information from our systems. Job status, proof status, changes in instructions and even changes in due dates should be entered into MISs. Reports can be pulled on demand or printed and distributed prior to the production meeting so the discussion focuses on key items rather than past history or information of little general interest.

　　Donald H. Goldman recently joined Prism-USA (Plymouth, MA) as senior project manager. Prior to that, he was the director of consulting services with the On Demand Printing & Publishing/Document Outsourcing and Corporate Consulting Group Service at CAP Ventures (Norwell, MA). Goldman also was the CIO/COO of Master Graphics, a multiplant printing-industry conglomerate. He has pioneered developments and educated the industry in many areas, including computer-aided estimating, digital prepress and workflow management.

[时间：2003-11-17  作者：Bisenet  来源：Bisenet]