14 Steps to Consider When Implementing MES
How do you implement your manufacturing execution system (MES) solution? Are you relying on a manual process involving spreadsheets and time-consuming manual methods that could be more efficient? All manufacturing production processes rely on some form of an MES solution to drive their performance, to manage their data. Some processes are outdated and inefficiently running, contributing to your company’s rework, scrap, or even adversely affecting your bottom line return on investment (ROI).
Are you ready to learn about the 14 steps we here at Sepasoft recommend you consider when implementing your MES solution? If so, read on!
We’ve done years of MES implementations and we’re here to share with you our recommendations for not only successful but efficient implementation solutions. Our implementations focus on best practices and lean manufacturing…calling upon the ISA-95 standard, Six Sigma, and other smart manufacturing concepts to help boost your quality of data along with a better way to efficiently manage your data!
As we dive into the steps, we will be outlining for you the following 14 steps:
1. Recognize the Need
The truth is, all companies have an MES system. It might be a completely manual process that uses paper, spreadsheets, manual communication, and etc. It might also be that a company has a partial or old system(s) that are limping along just enough to keep the company running. And there is a company that is so small that all details can be tracked in a single person’s mind. Which, there is absolutely nothing wrong with that. The fact that you are reading this paper tells me that there are shortcomings with your current MES processes. If you are pitching the need to upgrade your MES system, then identifying the top issues that are causing inefficiencies or maybe even chaos within your company is the first step.
2. Selling the Need
You may not think of yourself as a salesperson but this step must happen in order for an MES implementation to be successful. Without a question, there will be naysayers because some people don’t really like change. To get the naysayers on board with an MES system, they will have to be sold at some point along the process. Sometimes this will happen early on and sometimes it will happen after the start using it and see that it is not such a scary change.
It is also very common for there to be competing solutions being considered within your organization. This can become political and selling against competing solutions.
Behavior change is hard and a successful MES implementation requires pushing the naysayers, and those who are resistant to change, to move out of their comfort zone into the new behavior that is needed. Upper management being on board provides the extra push in the direction that the company will be implementing a new MES system so get on board or find the door.
Upper management is also needed to make the case to get funding approval.
It is vital to consider each persona and find out what problems or challenges each persona are experiencing and build into the MES system tools that help solve the problems. This will help with the ownership.
If the implementation is large in scope, it is important to break it up into phases. Each phase should include who the customer is, the problems that are to be solved, boundaries, and the success criteria. If you are an integrator reading this paper, the customer doesn’t mean the company that you have a contract with. The customer is who will be using or interacting with the MES system. For example: line operators, shift supervisors, upper management, etc.
The ideal MES system can write data to and read data from machine controllers. It is not a hard-fast requirement but if it doesn’t exist, then it will increase the task that your operators will have to perform. In addition, a manual system will not be as accurate due to data entry errors. Take the scenario of having the operator enter in downtime manually, the line goes down and the operator focuses on their first priority of resolving the cause. Now the line is running again and they go to enter the downtime event and they estimate the duration. An automatic system will be accurate and track the downtime event to the second. The operator can always add more details that will be helpful when closing the loop and improving efficiency.
To take advantage of sharing data with machine and process controllers requires the MES system to communicate with them. If you are implementing an MES system for a plant that has been around for a while, communications to the machine controllers may not exist. In some cases, this might require replacing controllers because they don’t have communications features. Next, the network must be present at each machine controller.
In addition to connectivity to PLC and other devices on the plant floor, operator terminals, barcode scanners, overhead displays and the like might also have to be purchased and installed.
Depending on the interaction with production, servers may need to on-premise for speed reasons or the risk production is stopped during WAN interruptions. This is commonly the case when track and trace systems are implemented. If the production rate is high and a 500mS delay will cause a brief pause of a machine while it waits for a response to verify the selected lot, then an on-premise server of some sort is needed. Likewise, if there is an extended WAN interruption and loss of production cannot be tolerated, then an on-premise server is needed. I have no question that in the future, WAN connections will be reliable and on-premise servers will not be required. In the meantime, on-premise or a hybrid architecture will have to be used. The hybrid solution has a smaller server that only maintains a limited window of future and past production data. The complete production data is maintained in the cloud or private data center.
The hybrid solution also helps reduce the ongoing cost to maintain backup or redundant, databases, backups, etc. The reason for this is that the on-premise server can be restored from the cloud servers because it is only limited data needed for production.
6. So Many Choices
There are a variety of MES systems out that range from canned for specific industries to configurable to any industry. Today, we are seeing more cloud options that are very enticing but it is important to not get emotionally hooked on an option before documenting some basic requirements first. These don’t need to be the full MES system requirements, but it is something you can share with vendors to see how they recommend fulfilling your basic requirements.
For example, your basic requirements should include production rates, estimated counts of resources like material items, disaster recovery requirements, level of service, handling of WAN outages, and more. Make these basic requirements relevant to what is important and the noise to signal ratio low. I have seen requirements that say something like must support reporting. The problem with this is that all systems probably support reporting and it will not help to narrow down the best option for your company.
MES systems are challenging to determine the ROI for because the return is hard to estimate. Answering the question of how much, if any, production staff can be reduced, how much quality will be increased, and how much will efficiency increase.
If an MES solution is implemented correctly, there will be savings. The real question is will there be enough savings to pay for the project? Every company that I have talked to after they implemented an MES system has stories of how they discovered benefits they never imagined at the start of the implementation.
The best approach to determine pre-implementation ROI is to look for inefficiencies such as reworked or scraped product, production interruptions caused by lack of coordination, etc. There is a very high probability that inefficiencies like these can be resolved by installing an MES system and an ROI can be calculated for them in advance for them.
Once you have the first facility or area within a facility complete, closely monitor and record the KPIs so that they can be used for calculating ROI on future rollouts. This will probably include the improvements that were never imagined.
Create a flowchart that aligns with this document that shows how to implement using Sepasoft modules and resources that we have available.
Over the past several years I have heard many stories from companies about their MES implementations. Most succeed in some capacity but there are also failures. Sometimes multiple failures. Just about all of them had a rough road and a very few were implemented smoothly. Some were implemented by an Integrator and some by the company itself. Some used ISA-95 as a guide and some didn’t. Some went through an exhaustive upfront project definition phase and some used a more agile methodology. Some had a sponsor from upper management and some didn’t. Some had great project management and others had no or poor project management. What is the difference between successes and failures?
This paper details the different steps involved in implementing an MES solution. Not all of the steps are performed by all companies but I listed them and the benefits of each and where they are needed.
You will notice that there are many steps to implementing an MES solution that is not technical. Many times MES implementations focus is on the technical aspects, but it is my intent to expose the non-technical steps that directly impact the success of an MES implementation. I also intend on exposing more details about the technical aspects that impact success.
9. Project Management (Technical)
Far too often I have seen an engineer left to his own decisions when laying out the architecture of an MES system. Sometimes this works out and sometimes the engineer hasn’t had the proper training, overcomplicates the implementation, or doesn’t use techniques that can be easily rollout. This usually ends up in project delays and a system that is difficult to maintain after the initial implementation is complete.
Because MES implementations involve many departments, the need for a project manager that handles the coordination between all of the different entities.
10. Will it Work? (Non-Technical)
This step is often skipped but is very important to success. Anytime operators, supervisors, management, maintenance, and others experience incorrect data or bugs, it reduces the credibility of the system. It can even re-enforce the naysayers who said it would work and the sad part is they’ll gloat as they tell co-workers.
The most important method to keep credibility high is to fully test the system before others that are not part of the implementation team, start using the MES system. This can be done using the live system during production, but hold off on turning it over to production staff until all counters, mode, states, KPIs, reports, etc. have been fully tested.
11. Training (Non-Technical)
This item is commonly overlooked but has a significant impact on the overall success of an MES implementation. Operators need to be prepared to quickly handle the various production issues that arise to prevent the system installed to increase efficiency actually reducing efficiency. It is good to include operator training in advance of the startup but after the user interface is stable.
Training of maintenance staff on the basic operation, architectures, and data flow. When PLCs are changed by some that don’t understand the ramifications that it may have on the MES system or not understand how to handle unexpected conditions in data flow will reduce efficiency.
Other departments that might need training include IT, management, scheduling staff, etc. Identify those that will be using or supporting the MES system and be proactive in bringing them up to speed prior to going live.
Training is an excellent time to pre-test the MES system. Usually, if some time was overlooked or is not working correctly, it will appear during training. Also, it is a good time to see how people interact with the MES system so that improvements in usability can be made.
12. Pilot (Non-Technical)
Pilots are not really needed but there is a time when they are essential if you don’t want to be associated with a failure. There are a few important cases to have a pilot. In fact, if you don’t go through a pilot in these cases you are putting yourself, your company, and at risk of losing credibility and money.
If what is being implemented is technically challenging or hasn’t been done before, then a proof of concept must be done. Basically, you have a theory and it is just smart to test it before banking the farm on it.
If what is being implemented doesn’t fit in with the culture of your company, then a pilot to test if the behavior change is achievable. This is also an opportunity to be agile and improve your solution to better fit your processes, company culture, and etc. If you position it as an open pilot where ideas are welcome, then most will be focused on what will help make operations run smoother. This is a better approach than having such a big investment in the implementation that new ideas are rejected because they are simply too risky with development being so far down the road.
Pilots are valuable in collecting ROI data. The only way to accurately know what you can expect for an ROI is by doing a pilot. Again, this is an opportunity to use an agile approach to learn what will fit well with the company and provide the greatest returns.
13. Are We Done Yet? (Non-Technical)
Buying a car gets you nowhere, you have to drive it. The same is true with an MES system. Implementing an MES system is like buying a car and using it is like driving the car. So what does “using it” mean? Operators seeing their efficiency, quality, etc. is one obvious use of the MES solution. Determining root cause analysis, resolving top downtime reasons, catching quality issues early, reducing cycle times, and tracking material flow are just a few samples of using the MES system to complete the MES solution.
I used the term MES solution instead of MES system throughout this document. The reason is it requires much more than just the hardware and software. To be successful, teams should be organized to evaluate the data from the MES system and turn it into actions that will result in more efficiency. The MES solution requires technical systems, training, analysis, new procedures, behavior change, and more.
I bet there are many Enron employees that wished they looked at their 401K accounts and made the decision to sell before a majority of their assets vaporized. If you don’t analyze the data and take action, then what is the point? Just to say our company implemented an MES solution is not enough you have to live and breath MES day in and day out.
Implementing an MES system is analogous to installing a flywheel and with no effort to start spinning it. It requires effort to start it spinning and the more people you get trained and onboarded to help spin it, the more momentum and the greater the return.
Chances are your company is in constant change and as a result, your MES systems will be in constant change as well. New products are added, machines are changed or moved, new initiatives are implemented that all affect the MES solution. The support of MES system to accommodate these changes in the business should be planned for from the start.
The rollout phase is commonly overlooked but can determine the success or failure of an MES implementation. Consider the time required to roll out to each manufacturing facility. If your company has just a few production facilities, then taking 6 months to roll out to each new production facility will not have much of an impact on the success of the MES enterprise implementation. However, if you have 25 manufacturing facilities and each takes 6 months, then it will take over 12 years to rollout your MES implementation. This will have a major impact on the success of the MES enterprise implementation. Simply put, you have to have your act together to reduce the amount of time to roll out to each manufacturing facility.
The typical items that increase the rollout time include hard coding, funding, over complications, scope creeps accommodating each facility’s desires, etc. I used the term hard coding to represent a structure that requires a lot of manual code changes for each manufacturing facility. UDTs (User Defined dataType), templates, parameters, standards, etc. must be used to easily configure a system to the equipment model of each manufacturing facility. This becomes even more important rolling out updates after the initial rollout.
Many manufacturing facilities have to have infrastructure added which is time-consuming but with the MES products available today, there is no reason for the configuration of the software for a facility to take more than a week.
Instead of making your initial rollout include everything that is asked for, start with a Minimum Viable Product (MVP) and rollout updates including new features over time. Taking this approach will greatly reduce scope creep and reduce the risk of a filed rollout.