Any values reported by or measured at a manufacturing machine are machine knowledge. This knowledge may describe what the machine is doing, particular course of variables, output knowledge, high quality knowledge, or knowledge in regards to the surroundings inside the machine or the situation of the machine itself. Machine knowledge assortment additionally entails knowledge held on different programs used to handle the plant.
Forms of Machine Knowledge
Machines outfitted with PLCs can accumulate knowledge about processing situations or the work carried out. Machines can be instrumented to assemble knowledge about their situation. There are a number of alerts and data a machine knowledge assortment system can collect, equivalent to the next.
Sensor knowledge
Among the industrial machine sensors used for course of management can additionally seize machine well being knowledge. Examples are temperature, vibration, present draw, and oil situation. More and more, machine builders are utilizing “sensible” sensors the place edge computing capabilities filter or course of knowledge earlier than transmitting it over a community.
Log knowledge
Any such knowledge quantifies what the machine has completed and is usually retained on a PLC or in PCs. It consists of machine manufacturing knowledge equivalent to portions processed, cycle instances achieved, yields or measurements, and different inspection outcomes. It will probably additionally cowl operational data equivalent to run time and downtime.
Community knowledge
Some fashionable gear comes with IoT sensors put in that may ship knowledge to different programs such because the CMMS. In different circumstances, producers retrofit machines with IoT gadgets. Knowledge is distributed over wired Ethernet, Wi-Fi, Bluetooth, and even mobile networks when longer vary is required.
Many course of crops function SCADA (supervisory management and knowledge acquisition) programs the place instrumented gear shares standing data with a central controller. This may very well be course of parameters equivalent to temperature or gear situation, e.g., whether or not a valve is open or closed.
Why Is Machine Knowledge Assortment Essential for Producers?
Manufacturing machine knowledge offers visibility into what’s occurring within the manufacturing facility and may present insights into gear operation and efficiency. Makes use of embrace the next:
- Predictive upkeep: Machine situation knowledge can drive a upkeep technique the place work is finished simply earlier than it turns into essential. This reduces upkeep prices whereas additionally enhancing machine availability.
- Yield enchancment: Course of knowledge reveals sources of variation that end in merchandise being outdoors of specs. This could drive enchancment efforts to search out and cut back waste.
- Growing utilization: Changing knowledge on whether or not a machine is working, being modified over, ready, and even not powered up right into a snapshot or cumulative report supplies the visibility to deal with causes of non-productive time.
- Course of optimization: Log and sensor knowledge can reveal patterns and correlations that allow high quality and productiveness enhancements. Optimizing drying instances and coating thicknesses could be an instance.
Methods for Gathering Machine Knowledge
Probably the most profitable implementations are the place measurable enhancements are achieved rapidly. This could imply specializing in downtime discount, upkeep value financial savings, or bottleneck utilization enchancment.
Begin by deciding what knowledge to gather, then contemplate the right way to acquire it. This implies establishing machine knowledge assortment objectives earlier than selecting and putting in sensors and shopping for analytics software program.
Having recognized a goal, an audit will present what knowledge is already obtainable. (It could be on PLCs, within the High quality Administration System [QMS], or within the ERP system.) This can even reveal gaps the place knowledge is required however isn’t obtainable.
For knowledge that’s presently unused, examine methods of getting it to a central location for storage and evaluation. This may occasionally imply including Ethernet connections to machines and establishing communication strategies. Newer machines could have such skills built-in however older gear may have specialist gadgets added.
The place knowledge gaps exist, examine applicable sensor applied sciences. Present draw, for instance, can point out whether or not a machine is working, idle, or not powered-up. For predictive upkeep extra temperature knowledge, together with vibration and circulate measurement, are sometimes key parts in increase an image of machine well being.
Implement the sensors and the connectivity options chosen. Use analytics instruments to interpret the information, and search for patterns that present new insights into the explanations numerous occasions or situations happen. Establish key metrics and arrange dashboards to make these seen.
Key Metrics to Observe
Collections of machine knowledge don’t have any worth till they’re changed into actionable data. One method is to calculate and share or show metrics that allow folks see what’s occurring and the place they need to make enhancements.
Among the critically essential metrics are as follows.
Uptime and downtime
Realizing whether or not machines are working, or for the way most of the obtainable hours they’re working, is vital to discovering productiveness enhancements and figuring out upkeep effectiveness. That is particularly essential within the bottleneck course of.
OEE
OEE (Total Tools Effectiveness) is used to evaluate asset utilization and determine the place to focus enchancment efforts. Knowledge from PLCs and sensors like present displays can present among the numbers for the OEE calculation.
First cross yield
This exhibits how a lot capability is being wasted by merchandise not assembly specs. It will probably typically be obtained from machine logs. Machine upkeep can typically cut back the variability that causes a low first-pass yield.
MTBF
MTBF (Imply Time Earlier than Failure) is an indicator of upkeep effectiveness. If predictive upkeep is being utilized successfully, machines will run for a very long time with out breaking down.
Power prices
Producers can measure the consumption of energy-intensive processes by putting in present or different sorts of meters. This can steer administration in direction of the largest value discount targets.
Manufacturing quantity
Extra granular data on order standing improves visibility, enabling higher schedule adherence, increased machine utilization, and extra correct forecasting. Whereas the manufacturing execution system (MES) often tracks orders, capturing knowledge immediately from the machines offers a direct indication of manufacturing fee and yields knowledge like cycle instances that can be utilized to enhance manufacturing planning.
Wanting Forward: Future Tendencies in Machine Knowledge Assortment
The variety of knowledge assortment instruments obtainable to managers continues to develop whereas their prices fall and capabilities improve. IoT sensors are enabling far higher volumes and charges of information seize than has been the case earlier than.
This knowledge will more and more feed right into a “digital thread” linking machines to the CMMS, components inventories, and manufacturing schedules, all to scale back waste and enhance effectivity.
An rising problem is the right way to take care of the amount of information that can be captured. A predictive upkeep system for instance could embody dozens of sensors put in on a whole bunch of machines.
All through manufacturing, AI will develop into important. AI in predictive upkeep, for instance, will determine patterns and developments in machine well being, enabling it to suggest and schedule service and even order alternative components.
The rising availability of machine knowledge offers producers new methods to search out and make operational enhancements. Elevated visibility and improved forecasting capabilities will result in issues being addressed on the first indication and earlier than they develop into expensive and disruptive.