Decoding Production Automation: Exploring Types and Technologies
Michael Chu
Product Management | Product Marketing | Product Development | Program Management | Process Improvement
In my previous blogs, I have explored various aspects of industrial manufacturing, such as machine vision, additive manufacturing and AI.
Today, I want to step back, take a tops-down view of manufacturing and talk about the systems that are important as we make the transition to a more intelligent and automated factory.
Specifically, we are going to look at one of the most important phases of automated manufacturing, production automation.
Production Automation and the Manufacturing Flow
If you think about it, almost all manufacturing, however diverse and complex, can be broken up 3 main functions:
In this context, production automation is all about the "Make It" step.
Depending on what you are making, you will need different types of systems to get it done.
Each of these production goals will have a different production automation system that is best suited to successfully making the product.
Most production automation systems can be categorized as one of the following 4 types of systems:
Distributed Control Systems (DCS)
As the name suggests, Distributed Control Systems (DCS) distribute the compute all throughout the factory or production line. While there may be centralized monitoring or a console to check on the status of the production, there is no central operator control.
One of the key advantages of DCS is that it increases the reliability of the entire production line.
DCS systems tend to be used in large process manufacturing plants where reliability and security is very important. Some examples of products where DCS might be used include: chemical plants, oil refineries, pulp and paper mills, boiler controls, power plants, food processing and pharmaceutical manufacturing.
Each of the nodes in a DCS tends to be lower power and ruggedized.
In certain circumstances, we may see a move towards centralizing the compute portion of DCS into an on-premises rack-mounted server. This would useful in situations where the entire production pipeline is geographically located close together and the factory owner wants to reduce factory footprint or make managing the nodes (now virtual) easier.
Machine Controllers
领英推荐
If you have a repetitive task that needs to be done, such as packaging your product into boxes, wrapping wires on a battery or 3D printing a custom part, you need a Machine Controller. Machine Controllers continuously take in input, figure out what control parameters need to set to using application-specific logic, make the action happen and repeat until the product is made.
The controls may not always be simple. More complex controls can be accomplished with Computer Numerical Control (CNC) where the control movement is handled through computer software on a CPU embedded inside the tool. CNC is commonly used in machining metal and plastic parts using tools such as routers, drills, water jets, lasers and more.
Machine Controllers usually have 3 distinct components, the logic that takes in the inputs and decides what to do, the motion control that actually performs the action and the human interface (HMI) which lets an operator start and stop the machine and check on the progress of the production.
As compute power increases, we will start to see the separate components of machine controllers get increasingly combined onto a single device. Also, expect to see AI be incorporated more and more to make the machine controllers more intelligent and efficient.
Stationary Robotics
Stationary Robotics are robots that perform their tasks from a fixed location. They can be attached to any number of surfaces, including the floor, ceiling or a platform. We can think about them in terms of two types:
While they may seem like machine controllers, stationary robotics are usually more complex, requiring several degrees of freedom. Collaborative robots may also require machine vision and AI to ensure safe operations.
Machine Monitoring and Management
Finally, while maintaining the systems in a factory may not seem relevant to production automation, it may be one of the most important system of all. One of the most important metrics to a factory owner is factory uptime. The more time that factory production lines are up, the more they can produce, which means increased revenue. When equipment breaks unexpectedly or machines get out of adjustment, the end production line suffers.
Using sensors to collect data from key pieces of equipment in the production line allows monitoring and management systems to provide real-time visibility into the operations and help drive improvements in the overall efficiency of the factory.
Some factories are even using machine monitoring systems to help reduce unscheduled equipment downtime. Using AI, they monitoring systems can preemptively predict machine failures allowing the owners to schedule a fix at a time convenient for them.
Next Time...
With so many different types of production automation systems, no single provider will be able to supply all of the systems needed in a factory.
In the next blog, we will explore some of the current and upcoming standards that will allow all of these systems to interoperate and work together seamlessly.
Credits/Acknowledgements: