OEE, Overall Equipment Effectiveness, in a single number, tells you how much waste there is on your shop floor. It is the ratio of what you have actually produced to what you could have theoretically produced. It tells you how efficiently your equipment and investment are being used. OEE allows you to measure and reduce waste (of available time, machine capacity, raw material) on the shop floor. It is a single number that you can communicate to, and is understood by, everybody from the machine operator to the President. It can be tracked month-on-month, improved upon.
OEE calculation formula
OEE = Availability x Performance x Quality
The components A, P and Q in the formula of OEE can be summed up representing:
- What duration did the machine run versus the time that was available ?
- How efficiently did it run ?
- How many good parts did it produce while it was running ?
Availability: How long was the machine available to run ? The ratio of the time that the machine was available, to the total time planned. The difference is because of down time caused by breakdowns, setup, no raw material, etc.
Performance: How efficiently is the machine running? The ratio of the number of parts produced to the number of parts that theoretically could have been produced in the time that the machine was running. The difference could be because of long part unload-load times, or cycle times than are longer than the standard cycle time. On a CNC machine the difference is because feed rate override or spindle speed override is not at 100%.
Quality: How many good parts were made ? The ratio of the number of parts that pass quality inspection to the total number of parts made. The difference is the number of parts rejected.
How does a Real Time OEE Monitoring System work ?
The system works on Industry 4.0 and has these components:
- Data capture from the machine in real time, the moment something happens. This is done electronically from the machine – signals from PLC, CNC, or retrofitted sensors. E.g., when the machine starts producing a part, the event data and time are noted down. Data capture can be done for any relevant event that occurs on the machine – cycle start, cycle end, spindle on/off, machine motions, tool change, coolant on/off, machine breakdown, etc.
- Send the event data to a computer that has real time monitoring software. The software analyzes the data, calculates OEE and various other parameters.
- Reporting the data to people, or sending it to other software, or controlling the machine based on logic conditions. Sending alerts on mobile phones if an issue requires immediate attention.
How does a Real Time OEE Monitoring System help ?
The typical traditional method of getting data from machines is one of these:
- Someone goes to each machine periodically (maybe every hour), asks the machine operator how many parts have been made, any downtime durations, and reasons for the downtimes. The person writes the data for each machine in a register. At the end of the shift or end of the production day (which is next morning), all the data is entered in an ERP or other software at one time.
- Machine operators write downtimes and downtime reasons in a register. At the end of the shift they take the register to the supervisor and give them all the data. The supervisor enters all the data in an ERP or other software before leaving for the day.
The first problem with both these methods is that all the data is based on the operator’s memory or truthfulness. OEE is critically dependent on this data: how many parts are made, how many parts are rejected, the downtime durations and the reasons for downtime. If this data is wrong, the OEE, Availability, Performance and Quality are going to be wrong. This is a problem of data inaccuracy.
The second problem is that if an issue requires immediate action (like abnormally low OEE caused by high rejections), decision makers in the organization do not know about it. They only come to know about the problem after a shift or a day, when it is too late to take action. This is a problem of data delay.
The longer the delay between updations, the more stale the data is, and the less useful in fixing problems. The longer you take to fix a production problem like a machine breakdown or high rejections, the costlier it gets. Updating every shift or every hour is useless. Updating in real-time is ideal, so people know about a problem the moment it happens. This is only possible with a real time OEE monitoring system.
Tracking the A (Availability) using a real time OEE monitoring system
The machine monitoring software uses special logic to determine downtime durations that could be any of these – power shut downs, no raw material, machine breakdown, setup times, no raw material, etc. Planned downtimes like meal breaks, time for autonomous maintenance, etc. are configured in advance for each machine. The monitoring system uses all this live and pre-configured data to determine Availability accurately.
It can also analyze data and show Pareto and pie charts of downtime reasons, downtime trends, big downtimes, etc. You can use these to focus on top downtime reasons and reduce them to improve Availability. It can send you alerts on your mobile phone for abnormally high downtimes, machine breakdown, etc.
Tracking the P (Performance) using a real time OEE monitoring system
The machine monitoring software uses cycle events captured from the machine automatically to determine how many parts have been produced at any given time. Standard cycle times and standard load/unload times are configured in advance for each part and operation. The monitoring system uses this live and pre-configured data to determine Performance accurately.
It can show reports of hourly production, details of each individual part cycle, planned vs. produced quantity, cycle times and load-unload times. You can use these to narrow down to any poor performance issues like poor operator skill level, misuse of feedrate override, etc. You can then fix the problems with appropriate solutions to improve Performance. It can send you alerts on your mobile phone for abnormally high cycle times, production target shortfall, etc.
Tracking the Q (Quality) using a real time OEE monitoring system
Rejections can be tracked through an operator input, or from an electronic signal from the machine. E.g., in an installation of LEANworx, a high pressure die casting (HPDC) machine is being monitored. A robot unloads the part from the machine and shows it to a CCD inspection camera. The camera accepts or rejects the part and signals the robot accordingly, which then puts the part in the accepted or rejected bin. This signal is read by the LEANworx IoT sensor hardware to track rejected parts.
Our thinking needs to change, from manual to real time OEE monitoring
Traditionally, the IT network in a manufacturing organization covers all the functions outside the shop floor – Stores, Finance, Accounts, HR, etc. Decision makers have accurate and timely date data in all these areas, and can take quick decisions. In this sea of IT, however, the shop floor is an island. The shop floor is where money is made. However, you’ll find hundreds of computers in departments supporting the shop floor but only a handful on the shop floor. We do not talk directly to the machines, and instead get data through a long and inaccurate process. OEE is as a result rarely accurate and in time.
A real time OEE monitoring system based on Industry 4.0 connects the office IT to the shop floor and to the machines. It enables decision makers to get accurate and timely data so that they can make accurate and timely decisions.