6 AI motion control applications to improve OEE
AI offers end-users the ability to analyze information collected from machines or manufacturing lines to optimize operational performance. Below, see six common motion control applications that can benefit from AI.
- Understand how AI can be used to optimize operational performance, efficiency and consistency for discrete pieces of equipment and manufacturing lines.
- Discover six specific AI motion control applications.
- Learn how AI toolboxes that organize and display data can suggest actionable decisions to end-users.
Artificial intelligence insights
Used effectively, AI can be a powerful tool for optimizing the efficiency and performance of manufacturing equipment.
Potential metrics AI can analyze include motor temperature, vibration, move profile information and more.
AI toolboxes can be employed to organize and visualize data so that actionale decisions can be made.
The use of artificial intelligence (AI) is becoming more and more common in industrial applications. The ability to collect information from a discrete machine or from an entire manufacturing line and use this information to optimize operational performance, efficiency and consistency can be extremely beneficial for manufacturing companies striving for higher overall equipment effectiveness (OEE). Used effectively, AI can be a very useful tool for unlocking the potential of manufacturing equipment.
Automation companies have built functionality into their products to allow the user to access all sorts of information from the automation products themselves and from the equipment they are automating. This information can be crucial in diagnosing issues that result in downtime while at the same time unlocking untapped machine output potential.
1. Motor temperature
Variances in the operating temperature of a servo motor can indicate changes in machine mechanics. Servo motors only draw as much current as is required to meet the application needs. Increases in servo motor temperature (which directly relates to the current draw) can most often be attributed to worn mechanical transmission components. This information is captured by the system controller and can be used in a preventative maintenance program where mechanical components get lubricated or replaced when their efficiency starts to drop.
2. Vibration detection
Some high-end servo products incorporate acceleration sensors into the control board of the servo motor’s encoder assembly. These types of sensors can detect vibrations seen at the motor during a normal machine cycle. The system controller can monitor changes in these vibrations over time and recommend corrective action that could be taken.
FFT (Fast Fourier Transform) analyzers in the system controller can provide a visual representation of what frequencies are present in a given set of machine data. These FFT “fingerprints” for each machine axis can be used to identify specific changes in machine mechanics over time.
3. Move profile information
Each servo axis on a machine has a unique move profile that is programmed for each application. The actual position of each servo axis can be tracked during its move profile and this data can be recorded at the time of machine commissioning. The actual position of each axis can then be tracked while the machine operates over time. When changes to the actual position of the axis during the move profile are detected, the system controller will identify specific differences (ie: overshoot, undershoot, higher deviation from commanded position, etc) and recommend appropriate action that can be taken. One example of an application attribute that could affect the actual position during a commanded move is a change in load inertia. If the load inertia changes even slightly over time, this will affect the overall tuning of the system and could contribute to significant changes in performance during the move. This issue could be remedied with adjustment of specific servo tuning parameters and/or rectification of the changes to the load inertia (ie: maybe there is a disturbance to the load being moved during operation).
4. Life monitors
Life monitors on critical wear components of automation equipment can be used to predict failures before they happen. Machine operation can be optimized by scheduling component replacement prior to a catastrophic failure that could shut down the manufacturing line. Motion control components typically come with their own life monitors as shown below.
A variety of other types of sensors can be added to a motion control system and monitored by the system controller. With a good set of AI algorithms, the system controller can take this data and predict future failure of specific machine components or just monitor overall machine performance.
5. Energy consumption
Electricity consumption monitors can highlight inefficiencies in individual machines or entire manufacturing lines. This information could be used to adjust machine cycles to minimize overall electricity consumption or to maximize energy consumption during certain times in the day when energy is less expensive.
6. System communication errors
Most sophisticated motion control systems allow for monitoring of low voltage supply power to each motor’s encoder. Spikes in this power supply due to system electrical noise can cause momentary loss of position data which can result in disturbances to the move profile. The system controller can detect when electrical noise is introduced to the system and how this noise affects operation. This information can be used to troubleshoot and correct the issue so the machine can run smoothly.
System controllers can also detect errors and loss of packets from the deterministic network being used (Such as EtherCAT, Mechatrolink, etc). Consistent network errors and loss of information can cause a variety of issues with manufacturing equipment. The data logging capability of the system controller can also be used for troubleshooting to correct these machine issues.
AI tools for organizing and communicating information
The availability of information is important, but having a set of tools that organize the information and suggest actions for improvement can be just as important. Some automation vendors have created customizable toolboxes that display the types of information described above as well as statistics related to the efficiency and performance of the machine or manufacturing line as a whole. These toolboxes can typically be viewed on the piece of equipment’s human-machine interface (HMI) or from anywhere in the world via a secure internet connection.
Tools such as this can be used for visualizing the operations of a single machine, a manufacturing line or an entire factory. They can provide asset management information, predictive maintenance schedules, alarm forwarding and data management. They can also exchange communications with external systems through OPC-UA or other common automation protocols. These tools have this base functionality but are also highly customizable.
The wealth of information available with today’s automation systems coupled with the highly engineered tools for organizing and distributing the information can be of enormous benefit to machine builders and machine users. Increases in machine throughput, higher part quality, reduction in machine downtime and increases in machine life are all achievable outcomes when machine and manufacturing line feedback is utilized effectively.
Scott Carlberg, product marketing manager, Yaskawa. Edited by David Miller, Content Manager, Control Engineering, CFE Media and Technology, firstname.lastname@example.org.
Keywords: AI, Motion Control
What other metrics could AI be used to analyze?