7 Myths about APS

One thing that catches the attention of academic and professional audiences is the lack of bibliography that specifically discusses APS. Basically, what you find online are materials with commercial bias (a bias we try to avoid in content like this on our blog, but well, here we are). The natural consequence of this is the limited knowledge the broader public has on the subject, learning about APS in basically two ways: in practice, actually using and/or implementing one of these solutions at work with the help of a specialist consultancy; or by word of mouth, talking with people in the first group or even attending a talk at some event. And that's how myths arise. Let's look at a few:1. With APS, I'll stop being late on my orders and creating stockouts.Contrary to the joke that teaches us to put 50 clowns inside a taxi, a factory doesn't have infinite capacity. Possibly APS's big trick is really working with finite capacity — that is, considering the limits and constraints of a production system to evaluate what will be possible to produce within the desired deadlines, or not. At the same time, APS will try to do this in an optimized way within the chosen heuristics, which really does generate more efficiency and allows doing more with the same (or with less).However, that last advantage makes some people think delays will simply disappear, as will stockouts for those who produce to stock. In some cases there's clearly a lack of capacity and, even by optimizing the factory's schedule, it won't be enough to meet demand. In those cases, APS will end up being useful for simulating capacity increments in specific sectors and supporting Capacity Planning so that, in the medium/long term, delays and stockouts are really eliminated.2. APS will schedule my factory by itself every time there's some unexpected issue during the day.The scheduling process within an APS software is constantly improving. It starts as a practical, effective process for PPCP to gain agility and good results from day one and gradually evolves with automations that only emerge from continuous use of the solution. At the limit, we can reach the point of generating a fully autonomous schedule that doesn't depend on human intervention. However, this is a long road and, even so, it's a discreet, bounded process.In the vast majority of cases, the APS user runs scheduling one or two times a day, and when there's a big unexpected event at the factory — a machine breakdown or a parachute order. It's even important not to reschedule constantly minute by minute, otherwise the schedule becomes, as we call it, very “nervous,” changing sequence at every moment. In that sense, it's good to have a frozen scheduling horizon and try to avoid continuous changes to it. At the end of the day, if there's some short-term spot issue on the shop floor, the one solving it will be the factory itself.3. APS only works if we have 100% of engineering master data up to date and correct.This belief is possibly the most recurring one and will never be true. No industry will always have 100% of its master data correct and up to date (here we mean engineering master data such as manufacturing routings with process times and resource alternatives, as well as product structures). Since a factory is also in constant change, even a great study can be outdated after just 3 months.To minimize the impacts of these imperfections intrinsic to reality, the ideal is to control schedule adherence — how much of what's being scheduled is actually being executed. If that adherence is low, the main reasons can be poorly designed scheduling rules, resistance on the shop floor to follow the schedule due to power disputes, incoherent production performance indicators, inefficiencies, or times and resource groups with master-data problems. Isolating the other reasons, you can identify where your master data may be most problematic and form task forces to fix it.Lastly, APS use is also very effective for identifying problems because it's very visual. If, for example, a production time is much higher than it should be, the Gantt chart itself will flag the problem on the spot, showing an operation longer than the others.[caption id="attachment_2065" align="aligncenter" width="522"]

If you depend on all master data before starting, you probably will never start.[/caption]4. APS will help me analyze my historical efficiency (OEE) and diagnose what problems I have in production.The word “scheduling” already implies the future. An APS system is focused on scheduling orders/demand in the future. That is, in this process, the past is not the protagonist. Of course it can help — especially the short-term past. NEO uses short-term shop-floor data to generate schedule adherence views (the KPI mentioned in the previous item), which is extremely useful.But you can't expect APS to store years of history and bring detailed analysis of it. Technically it's totally possible, but that's already a different solution scope, much more aligned with Production Control and MES (Manufacturing Execution Systems).5. APS only works if there's a Production Control (MES) solution in place. An MES is important for a real-time production overview (or as updated as possible), but, as in myth 2, scheduling is usually run one or two times a day. So if the key-user schedules at 8 a.m. and 1 p.m., what's really needed is for the scenario of what was produced up to that moment to be updated — not that everything is updated at every moment. In that way, postings at specific times of each shift, or postings per movement unit (pallet, coil, etc.), are already enough for good detailed scheduling in APS.6. APS isn't useful for automotive assembly lines.An automotive assembly line really isn't the traditional use of APS, but there are advantages that can be very positive with it. There won't be very complex synchronization between resources and operations caused by cross-flows in production that occur in other segments, but there are critical material synchronization issues. And APS handles this synchronization and helps all the processes adjacent to the assembly lines serve them as well as possible — identifying when inputs are needed at each line operation, as well as when line changes due to labor constraints are essential (which can cause line stoppages and are handled in APS).In the end, due to the criticality of keeping the “factory drum” always efficient, APS use focused on material synchronization and constraint control is very useful.[caption id="attachment_2064" align="aligncenter" width="610"]

Automotive line[/caption]7. I don't need APS because we've implemented the Lean Manufacturing System.Surprisingly, concepts that should add synergistically are often confused and seen as conflicting. Beyond the fact that Lean Manufacturing System is more than a methodology — it's a philosophy — it can heavily collaborate with APS and vice versa. The myth 6 example itself shows this, putting APS in an environment that traditionally breathes Lean. At the same time, the adherence analyses we saw in myth 3 can be raw material for organizing a Kaizen Blitz; Quick Tool Changeover (SMED) can constantly update the setup times used by APS, and so on.At the same time, Lean alone doesn't address the extremely dynamic bottlenecks that are increasingly frequent in productive environments where SKU volume and product diversification rise sharply, causing one sector to get overloaded and another idle depending on the mix to produce. That dynamism requires a quick, intelligent response that only a specialist system can provide — with heuristics and a simulation power capable of adapting to each situation and moment at the factory. Enter APS.Of course, these are only some of the myths (yes, there are many more!), and the “believers” in them don't always defend them tooth and nail. But we notice, implicitly or explicitly, these are questions that haunt the minds of those who study or practice APS. Reflection is worth it to understand what not to expect and what not to fear from it.Have you heard any other myths? Share them with us here![noptin-form id=2822]