In the aerospace industry we see black boxes used to save all sorts of pertinent data and even conversations up to the time of the crash. Dash cams in cars record data also constantly but overwrite the info unless pertinent. In the case of manufacturing and tooling crashes, the internet of things is the black box.
When used properly the tooling crashes can be used to reproduce the information leading up to the crash. I used to have a service manager who said, if you want to know what happened in a crash on the production floor, look in the garbage or recycling. After a crash occurs the suspect part gets thrown out (perhaps to hide the evidence).
The Internet of Things (IoT) and systems like Linknet can monitor machines status and the bypass of safety intended monitoring. In the past you can typically walk through a machine crash. Joseph Zulick is a writer and manager at MRO Electric and Supply.
First, there is an issue with the set-up, either because the wrong data, job number, call data, material or setting height or depth may be wrong. The result is the operation creates too much tonnage, motor draw, torque, etc. The second safety system then shows a bypass or jumper, so sensors used to confirm that everything is ok are not seen by the system.
The third system is then bypassed, tonnage monitor, torque limiter, etc. this allows the system that would have stopped the machine after the initial failure occurs, to stop, but instead it continues until a crash occurs. These crashes can cost tens of thousands in machinery and tooling damage and thousands per hour in lost time of production.
What do systems that record this data tell us? In many of the systems you record the job data. Then an alarm is triggered in the IoT reporting of sensors being ignored or bypassed; it also records the operator and setup who performed the task. This can also be automated using RFID tags and pass codes.
This should be done to prevent non authorised personnel from accessing the system. Once the bypass of the safety or tooling protection system is bypassed the secondary prevention kicks in to lock the system if it senses excess tonnage, torque or amperage. This should prevent major machine damage. All of this data is recorded then transmitted to a system compiler or hub which communicates to the software what has happened.
Many new systems are working to be smarter than the average damage. In many systems, once you go to run mode, automatic or full production, the bypass is automatically turned off to prevent an accidental situation where it’s left on by a careless employee.
This prevents the second and third bypass mistakes and they can only occur during a setup or initial manual mode. This means it is left up to an operator to handle this aspect. The more automated you become, the less manual operator errors can occur but the more high-tech you need to become to allow for automation to handle the setup. This requires more programming and more Iot sensors.
This is a great step forward but doesn’t come for free. Every stage of automation carries a higher price tag. Much like the return on investment (ROI) on the low hanging fruit you get from typical lean improvements. The higher you go, the higher the cost and the lower the return. But, this is the area that can save on these hidden costs.
All of the data gathered can piece together the crash scene. You will find some of the causes but you will still need to dig to determine the root causes. The more sensors and Iot you have , the clearer the picture will be after the crash.
We all need to realise this is only half the solution, the real goal is to not have crashes and to use this data to prevent any further occurrences from impacting your production.
Acting on the IoT data is the real benefit. Gathering data is like taking a picture of the empty bank after the robbery! You want to catch the production robber red handed! Artificial intelligence (AI) is very beneficial in this area, in the background AI is running analysis on the machine activity via interface using hardware interfaces which pull in real time data and the AI is then determining the meaning of the data. The long-term goal is prevention and prediction of failure.
In the short-term when a crash occurs the data gathered from the sensors, programs and interfaces provides the pieces to determine the cause. Digging down to the root cause is vital, this may mean more specific sensors, more data analysis. If you’re not measuring it, you won’t be able to put the pieces together. Look at your IoT system, examine for holes? What root causes are you not analysing? Plan, do, check, act.
The author is Joseph Zulick is a manager at MRO Electric and Supply.