Math is brutal in manufacturing. You can run well at each individual process step, say 99% good, which sounds like a solid A. But then you have to multiply all the different steps together to get to a final product, and it only takes 25 parts or processes, each running at 99%, to net out to 78% (0.9925 = 0.778). Turns out that 99% good at an individual station on the plant floor is actually lousy quality.
One way to fight math is to bring in bigger, “badder” math. Machines already spew out data with each cycle, so there is a growing appetite to analyze these data streams right in-station and identify instantly whenever some variable (temperature, psi, torque, rpm, you name it) starts to drift towards its control limits. When this happens, you can have decision rules in place that cause the machine to shut down and wait for help, or, if it’s very clever, self-adjust to bring the variables back in line. If you’ve captured all the relevant variables, this could bring an individual station up to something approaching 99.99%. This is what an A looks like in a factory.
The next step is to use technology such as the Dell | SAP HANA solution to integrate various data streams for a given product, such as:
- Machine cycle data from across the stations
- ERP system data such as maintenance records
- Component quality shared from your supply base
HANA allows you to analyze a massive amount of data in real-time to systematically detect and analyze trends. When trends point to a strong prediction that a quality issue is about to occur, the teams on the floor can take immediate action to prevent the issue.
While effective, it is still a lot of hard work to control everything to such a high degree of perfection. One engineering principle I remember well is that tight tolerances cost money in factories, and that excellent engineering is where tolerances can be relaxed, but an outstanding final product is still the result.
The ability to make adjustments on the fly is very much a human capability, one that we use hundreds of time daily as we drive on roads, navigate a business meeting, or cook a meal. Giving our machines this ability through data analytics and communication makes a factory truly “smart” – and capable of producing A-level work, cycle after cycle.
‘Internet of Things’
Most manufacturers have been standing on the sidelines when it comes to the Internet of Things. Here are some reasons why they should get in the game.
- 1The ‘Internet of Things’ changes manufacturing
- 2How the ‘Internet of Things’ will change design
- 3Becoming customer-centric via ‘Internet of Things’
- 4Securing the ‘Internet of Things’
- 5How the ‘Internet of Things’ impacts the factory
- 6Future ‘Internet of Things’ — from device to sky