APC (Automated Process Control) is a concept invented under the direction of Prof. Maurice Pillet by the research laboratory SYMME at Savoie Mont Blanc University. Automated Process Control is an evolution of SPC (Statistical Process Control) and consists of using IT resources to control a production process in order to obtain better quality results than with traditional SPC.
Today, Automated Process Control is used industrially in a wide range of areas, and the benefits are clear to see. In the case of machine tools, Automated Process Control (APC) has made it possible to reduce changeover times by a factor of 3 or 4, and to divide the scrap rate by 10. For some time now, Automated Process Control has been extended to other types of process such as bending, injection, extrusion, etc., with equally impressive results.
Automated Process Control uses software available to the setter which calculates the corrections to be made to the process based on the measurement of one or more products. Automated Process Control is based on two fundamental concepts:
- Controlling complex processes by mastering the simultaneous correction of dozens of correctors
- Correct drift using a damping function
Managing complex processes
When monitoring complex processes that combine multiple, mutually dependent measurements, such as the machining of complex parts, machine set-up is often a difficult task. Take the following example:
We want to correct measurement A because, with a measurement of 10.2, it is out of tolerance. To correct it, you need to modify the stroke of tool C3, but doing so will modify dimension C, which is under control. Tool C2 must therefore also be corrected. However, doing so will modify measurement B etc.... The problem for the setter is how to correct all the tools C1, C2 and C3 to correct measurement A.
In this simple example, the set-up may still seem trivial, but in reality the parts are more complex than the example above, so the set-up process can take a considerable amount of time.
Automated Process Control (APC) automates settings
To resolve this setting difficulty, Automated Process Control software such as the Ellistat APC module automatically calculates the setting for each tool.
The set-up process is automated using a configuration file, which can be built either:
- Based on the unit impact of each tool
- From a file of three-dimensional measurements of the part
1. Based on the unit impact of each tool:
In this case, configuring the Ellistat APC module involves calculating the impact of modifying each trimmer. The question to ask is simple: If I modify this offset, which characteristics will be modified?
Once all the unit impacts have been calculated, which only needs to be done once during the industrialisation of the product, the Ellistat algorithms can automatically calculate the corrections to be made for each setting. In the previous example, the result would be :
C1: -0.3 C2: -0.3 C3: -0.2
2. From the part's three-dimensional measurement file
In this case, configuration of the Ellistat APC module is even simpler, since all that is required is to import a 3D measurement file and configure the types of offsets used by the machine. Ellistat's algorithms do the rest, automatically calculating the optimum offsets.
Whichever configuration is chosen, the setter no longer has to calculate the setting, as it is supplied automatically.
Implementing Automated Process Control offers a number of benefits compared with traditional process control:
- Allows you to monitor dozens or even hundreds of characteristics.
- Automatically calculates the relationships between several correctors.
- No calculation or sign errors.
- Suggests the most likely setting and standardises the way in which settings are made between teams.
Correcting drift
The limits of SPC
When a process is controlled using the SPC, control charts are used. Control is based on control limits, which symbolise the normal operation of the process. When a measurement is within the control limits, there is nothing to do - the process is under control. Conversely, when a measurement is outside the limits, this means that the process has drifted and action needs to be taken, i.e. the process needs to be adjusted.
The problem with the SPC is that it only allows you to react when the process has already drifted significantly. In the following example, the process has drifted but only half of the measurements will lead to a result outside the control limits and therefore to a reaction.
The damping functions of Automated Process Control (APC)
The use of a computer tool to calculate the correction to be made enables us to be much more precise than a control limit which only contains the inside/outside information. In Automated Process Control, the process is corrected at each measurement to limit process drift.
The use of APC allows much better results to be obtained than with the traditional SPC. The examples below show the use of the SPC and the Ellistat APC module on a process with drifts.
Monitoring a process using traditional SPC :
Follow-up of the same process using Ellistat's APC software with Automated Process Control :
Conclusion
Automated Process Control makes it possible to automate the management of production processes, enabling previously unattainable quality results to be achieved and significantly increasing machine utilisation rates by minimising downtime due to adjustments.
Ellistat's APC module is the first software to fully implement the principles of Automated Process Control. Its application is simple and transforms the company's productivity and quality results.
Whether it's monitoring a process with drifts or controlling a complex process, the algorithms offered by Automated Process Control simplify the work of the setter and enable a whole new vision of process control to be imagined.
The benefits of implementing Automated Process Control are manifold:
- Number of adjustment parts reduced to 1 : Automated Process Control minimises the number of adjustment parts. The scrap rate due to adjustment parts (generally from 1 to 3%) is on average divided by 3.
- Minimum set-up time : By automating set-up, you only need one part to set up a machine. Set-up time is reduced to an absolute minimum, increasing machine uptime. For example, a gearbox casing that took half a day to set up before Automated Process Control was introduced can now be set up in 30 minutes (production cycle time).
- Maximised quality indicators : Automated Process Control enables the number of out-of-tolerance parts to be reduced considerably compared with conventional control. Experience shows that the rate of rejects due to out-of-tolerance parts is considerably reduced. Automated Process Control makes it possible to obtain Ppk quality indicators higher than 1.33 for all characteristics.
- One-day setter training: Applying Automated Process Control is particularly simple. It only takes a day to train a setter to use it.
