It should come as no surprise that most manufacturing companies are using Excel and paper-based systems to monitor production and get visibility into the factory floor. It kind of works if there are strict, buttoned-up systems in place. If there aren’t, these systems fail to provide the visibility most manufacturers need. Excel and paper-based systems will only take manufacturers so far, and then, it’s time for an automated system.
Don’t believe the hype that you need terabytes of data, an army of data scientists, or AI to replace Excel, paper, or whiteboards on the floor with automatic data collection. It’s so much easier than you were led to believe. The big kicker? Most manufacturers believe they lag behind their competitors if they’re still using Excel and paper-based systems, but that’s not true either.
The real truth of the matter is automatic data collection and greater visibility into the plant can start with just one single data count.
Where Manual Reporting Goes Wrong
As the manufacturing industry embraces the smart factory mindset, the first step begins with replacing manual reporting. There are a lot of issues with Excel and paper-based systems. It’s error-prone, and data entered into a manual system is subjective. An operator may record downtime as 6-minutes, but in reality, it was 12. Accuracy and consistency don’t exist. Rolling up data and looking at it over weeks, months, or even years is difficult. Sure, there might have an Excel wizard on the team that can build macros and gets the data needed, but the second the COO asks what happened during a particular week 8 months ago, finding the answers will require digging through tons and tons of data in spreadsheets or paper reports stored in a file cabinet. The problems with manual data collection go on and on.
The point is that data is not easily accessed in Excel and on paper, and it sure doesn’t provide great insight when it is accessible. There isn’t good, real-time visibility, and understanding what happened is only possible after the reports are complete. It’s quite the headache that no amount of ibuprofen can solve. Yet, a single data can wipe away that tension.
Automatic Data Collection Begins with a Single Data Point
I like to refer to my favorite quote spoken with wisdom by Lao Tzu, “a thousand-mile journey starts with a single step.” In that fashion, the smart factory journey begins with a single data point.
Implementing a smart factory system is relatively easy and cost-effective, providing lots of visibility with very little work… and that all starts with a count.
There have been systems around that have done this for a long time, but things have evolved since then. Starting with a single data point is the first step, but then you grow out of it. That’s where software-like manufacturing productivity comes in as it’s designed to grow with the manufacturer. It doesn’t stop at a single data point – it expands on the insights gained until all of the problems are solved and the plant is running efficiently.
A Single Count Yields Endless Possibilities
Believe it or not, there are a lot of things made possible with just a count – measuring throughput, tracking downtime, understanding performance, evaluating schedule attainment, and it doesn’t stop there. Through a single count, data is collected and decisions can be made to figure out how to improve, one step at a time.
A great example is downtime. The machine tracks each product with a click. If the machine stops clicking, the machine is down, and that is recorded. This tracking method, versus manual paper collection or Excel, is substantially more accurate because the machine is doing it, rather than a human.
With that data, other insights are available – tracking against the schedule in real-time, understanding if the plant is having a good or bad day, knowing if due dates on job production orders are met and if not, why not, seeing where bottlenecks are, and above all, knowing all of this data now, instead of later when the reports are finalized.
How to Collect Data
Collecting a single data point is relatively easy. It can be done in 4 ways:
- PLC or Controller: This method is surprisingly very easy. 20-30% of the machines in factories have PLCs or controllers that can be networked and provide data. This doesn’t require programming, and manufacturers don’t have to change anything about the machine itself. Systems like Mingo can read the data directly from the registers and turn it into something useful.
- Existing Sensor: If a PLC doesn’t exist or doesn’t have the data needed, data can be collected from an existing sensor, whether that’s a relay that exists on the machine, a photo-eye, proximity switch, or any other type of sensor. Productivity systems can piggyback off the sensor and grab the data needed.
- New Sensor: If the machine doesn’t have a sensor that can be used or there are complications, adding a sensor is an easy fix. There are all different kinds of sensors that can be used to count a single data point.
- Manual Data Collection via Tablet: If there’s no way to collect data automatically, using a tablet to record data manually is the fourth option. Rather than writing down data on a whiteboard or clipboard, an operator is typing it into a system. The benefit is a centralized, consistent system where no one has to compile and make sense of the data – all of the data is in the same system.
Versatech is an example of a company that employed manual data collection methods in addition to automated solutions and saw significant results. Regardless of the method, having a system that provides immediate visibility will make the plant more efficient and productive.
Getting the Most out of a Single Data Point
Collecting the data via any of the 4 methods above will yield significant results. What manufacturers do with that data makes all of the difference.
Show the Data Throughout the Plant
Once data is collected, don’t gatekeep it. Put up scoreboards throughout the plant, on the floor, in the breakroom, and in the office, to show different machines or lines, if they’re up or down, and provide reasons why. As employees are walking the floor, they’re able to look up and know exactly what’s going on. Not only does it provide visibility into what’s happening in real-time, but it gives supervisors the tools to quickly know how they can help and where.
Today, more than likely, manufacturers are collecting productivity information in an ERP or compiling data into Excel or paper-based reporting and sending it out. Regardless, these methods of reporting aren’t looked at during the shift and there’s no visibility into productivity until the end of the shift. But more often than not, there’s really no visibility until the end of the week when reports are complete.
With production monitoring software, reporting is automated and timely. Additionally, it allows employees to focus on other tasks such as solving problems versus just compiling data. We recently put a system into a brewery and the engineer said automating reporting saved him 2 hours per day. With 10 hours a week back in his schedule, he’s now able to work on and fix other, more important tasks.
The other big win? Automated emails for end of shift, end of day, and end of week reporting, mobile access for insight at a glance, and real-time alerts. Not only is it all automated, but no one has to do anything to make that level of visibility happen, other than entering data into the system.
The point here isn’t to showcase Mingo, but to show that little things like how a single data point can yield significant visibility and identify optimizations on current processes.
Proving the Success
Understanding what that data point can do is the tricky part. It takes knowing what problem needs to be solved and contextualizing the data needed to solve that problem, but gathering insights from Excel and paper-based reporting to make data-driven decisions is hard. With a production monitoring system, it can contextualize the data for you and look at trends over time. It’ll answer questions like, “How did we do the last time we ran that part? How did we do last week? Last quarter? Did the process improvement actually help? Did the process improvement stay fixed? How can we monitor it going forward?”
There are customers of ours who have answered these questions for themselves and as such, gained valuable insight into their plants that produced big wins. Take a look:
- Tacony: Reduced lead time 150%, reduced a 14-week backlog to gain $500k in unrealized revenue, and saw a 9x ROI
- Oral BioTech: Increased OEE 25%, reduced scrap by 99.8%, and created a culture of visibility and transparency
- H&T Waterbury: Reduced micro-stoppages by 71%, eliminated 18 hours of downtime, and drove continuous improvement projects
- Lyons Blow Molding: Eliminated manual reporting, increased accuracy, and mistake-proofed the plant
The common denominator between all four customers is a single data point. They started small with a data point, thought big, and move fast to achieve their goals. Start small, think big, and move fast is how we encourage people to implement these systems to get big returns and achieve goals. It all aligns with the power of a single data point. That’s the smallest thing you can do.
The first step is starting small enough to be successful. The goal is to guarantee a quick win to pick up momentum and keep the project going, get more improvements, drive visibility, and understand what’s happening in the plant.
Truly, time is more expensive than the money spent on these systems – it’s important to get it right the first time around and expand from there. To produce a successful implementation, it’s going to take a decent amount of time to get up and running, whether it’s one line, one machine, one cell, or 20. The more machine, more departments, and more lines set as a starting point, with a limited number of data points, the more likely chance of success. People are more involved, more resources are allocated, more focus is given, and all of this contributes to increased levels of success.
The reasoning for all of this comes back to efficiency. With a single machine, the odds of duplicate work are high. If one machine uses automated reporting, but all the others require an operator to record data via paper or Excel, they’re likely going to revert back to old habits, effectively making the production monitoring system useless. To be successful, the system has to be embedded in processes to deliver the kind of visibility expected.
The “think big” piece is understanding what’s next. What’s the end goal? What is the North Star? That will provide the direction needed to drive the project, with the caveat of doing it incrementally. Then, move fast enough to capitalize on the wins. Momentum is everything and getting quick wins every time something new is added will increase the chances of success. In a way, it’s similar to the scientific method. Establish a hypothesis, test the hypothesis, compile the results, and keep going and improving things over time until the hypothesis is proved. It’s really that simple.
Making it All Work
The best tip we can provide is to engage the operators on the floor. Data is needed from the operators to make all of this work, and quality data from them produces a quality outcome for everyone else. Likewise, bad data from the operator produces a bad outcome for everyone else. Talk to them about, “what’s in it for me?” The answer is simple – the data provides insight into the problems that plague operators’ shifts. It provides the context supervisors and managers need to identify and solve a problem, making each shift more seamless for operators.
Second, the other thing required when implementing a production monitoring system is consistency and effectiveness. Data needs to be entered regularly and it needs to be reacted to. If operators don’t feel supervisors and managers are looking at the data or reacting to it, they won’t understand why it’s being collected in the first place or support the effort. They won’t input data. They will game the system. Instead, use the data to solve problems and use the system as a tool to measure efficiency so the plant can improve overall.
Third, limit manual data collected as much as possible. Automated data collection leads to higher quality data, but there are circumstances where there’s no way around manual data collection. If that’s the case, align it with what’s already being done today. If product counts and downtime are collected once an hour manually, align the manual data collection to the automated data collection baseline, too. Require everyone to input the data at the same time, regardless of the method of recording. It establishes consistency and good habits.
Four, integrate as much as possible. It may be hard to do when just starting out, but try to avoid multiple systems that all work in a silo. A single, integrated system establishes processes that can easily work together, making everyone’s lives in the plants simpler.
Five, leadership has a massive role in the implementation process. If an executive isn’t already on board, recruit an executive to own the project so roadblocks can be easily cleared. Then, recruit a strong supervisor or line lead. Their role is to enforce the use of the system and to do so, they must believe in it. It’s much easier to get up and running, achieve a quick win, and replicate that process when a manager is behind it fully. That manager can also push the project forward and monitor it even if it’s struggling. Without strong leadership, the less likely the project will be successful.
Six, at the end of the day, the whole purpose of a production monitoring system is to turn data into information so problems can be fixed. In that vein, more data is not always better. Sometimes it’s just more data, and sometimes, it makes everything more confusing. Don’t try to boil the ocean.
Remember this, someone needs to own the project, someone needs to run the project, and someone needs to project manage the project. Identify those people so it will run smoothly.
It’s Easier Than You Think
Production monitoring implementations are nothing like stressful, expensive, and time-draining ERP implementations. They are the opposite of that. They’re more simple, more straightforward. Starting small with a single data point means implementation is quick, doesn’t take months and months of conference room pilots to figure out the process, and most importantly, provides a lot of information to get visibility into the plant. The goal is to learn, improve, learn, improve, and continue to do that over and over. That is the best way to get significant benefits and a return on investment.
Remember, it’s possible to take that single step of a thousand-step journey with the right mentality and resources available. To do that with production monitoring software, start small, think big, prove it works, move fast, and grow. That’s the goal of a single data point and real-time visibility into the plant.