Optimizing shop floor schedules has long been the Holy Grail of manufacturing. It’s so fundamental to manufacturing performance that it has been the subject of extensive academic research for at least three decades. Yet even with advancing computer technology and sophisticated scheduling algorithms, real world solutions have been hard to come by.
There are several reasons for this, among them:
- Lack of high quality inputs – It’s hard to know precisely what’s going on in production at any given time; the GIGO adage applies here (Garbage In, Garbage Out); schedules often fail when information is out-of-date or incorrect.
- Inability to collect all elements affecting the schedule – This problem is related to the previous one, viewed holistically. Many activities impact scheduling on the shop floor, such as equipment maintenance, people, machine availability, quality and materials, among other factors. Even the weather can play a part, impacting deliveries and shipments. Accurately capturing and applying all of this data has traditionally been a big hurdle.
- Limitation of “batch-based” optimization, especially when confronted with a disrupted event – This is simply a limitation of the basic assumption that a computer program has to take in all information, process them as a batch and generate meaningful output after a certain time period of processing, while real world events keep changing in an ever faster pace.
A New Era?
Despite these challenges, it appears the long-sought after goal of a process-optimized shop floor might now be within reach. A key reason is that technology has advanced to the point where processing is fast enough, algorithms are good enough and scheduling systems can now keep pace with the frequent changes that occur in manufacturing operations. Some of the key technologies are:
- Internet of Things – The technology of sensing and collecting operation activities data in real-time through tablets, cell phones, RFID, machine sensors, GPS, radars and the like has become ever more efficient.
- Advanced algorithms – Commercially available package software now has the ability to beat academically optimized algorithm in solving optimization challenges such as traveling salesmen problem or NP-hard problems. Moreover, agent-based technology is able to collaborate in real-time with human decision without falling in the trap of “batch-based” optimization.
That leaves one problem: getting the right information into these systems so that they can schedule accurately. And it looks like this final piece of the puzzle has fallen into place, in the form of the modern Manufacturing Operations Management (MOM) system capable of collecting high-quality data about events on the shop floor and beyond, in near real-time.
Advanced MOM Solutions Can Deliver Right Information to Right Person at Right Time
With a modern Manufacturing Operations Management (MOM) solution it’s now possible to capture nearly any relevant data point related to the shop floor—from machines and processes to people and suppliers—and feed that data directly into “live” optimization systems. This capability is starting to have a big impact.
For example, optimization tools traditionally have been applied on a weekly basis – that was the best they could do. Today, those manufacturers that have invested in a new enterprise MOM solution have found that they can apply scheduling optimization to real-time events within a single shift. When something happens to disrupt the schedule—say a quality defect or machine failure—these manufacturers can immediately generate new schedules that are optimized for all the various factors involved, taking into account not only resources but also customer expectations and value.
Some skeptics might still wonder: so what? Why not rely on gut instinct and experience, as shop floor managers have always done?
Well for one thing, even if you have a skillful manager who knows how to juggle all the orders and schedules coming down the line, you can’t easily replicate that expertise and put it on your other shop floors. But you can replicate automated scheduling optimization.
Just as importantly, even a “genius” manager on the shop floor is becoming hard pressed to deal with the complexities of modern manufacturing. Product lines are much more diversified, supply chains more flexible, customers more demanding, and competitive pressures to be lean and efficient more severe.
The goal in manufacturing these days is to synchronize production with demand to reduce waste and inventory and improve efficiency. Indeed, it’s becoming an imperative. Why optimize base on a weekly bucket when you can now do so on an hourly basis? With the right systems in place, the cost to respond to a schedule change is virtually nothing – but the opportunity cost of rigidity can be quite significant.
For manufacturers that have invested in upgrading their manufacturing operations management capabilities, the answer seems clear. I expect 2015 to be a good year for shop floor real-time optimization.
If you liked this article, here is another that might be of interest:
