Today in automation, predictive and preventive maintenance are constantly in the conversation. While the ideas may be clear, the implementation and advantages are not. Is this a fad that will fizzle out soon?
By Gerry Paci, Marketing Manager — Material Handling ⋅ Pepperl+Fuchs
For many years in the material handling industry, warehouse operators have strived to achieve minimal downtime and optimal operation. There are countless ideas and aspects for attaining this milestone within the warehouse.
In this era of intralogistics, the requisite for automation has become apparent. While some automate more than others, the necessity has proven clear. Managing automated equipment and potential downtime is crucial to production and intralogistics.
When a machine is down, the alarm will sound, reminding those on the floor that time and money are paramount. It is one thing to fix the issue quickly so that the process can continue, but understanding when this type of issue or downtime could occur again is the real prize. Moving from reactive to proactive maintenance in the warehouse can be the difference between a streamlined process and a chaotic event within intralogistics.
The industry has learned over the years that regularly scheduling maintenance helps minimize problems with automation. A planned approach involving inspecting and repairing equipment to prevent failures has proved successful. Based on history, the warehouse maintenance team may understand that a motor controlling the conveyor needs to be inspected once per quarter for potential loosening due to vibration. This tactic helps extend the equipment’s lifespan and identify and address problems before they cause downtime. Nonetheless, if the maintenance team had access to the motor’s vibration velocity and temperature over time, they could predict when a failure could occur, streamlining their maintenance process.
This information and data have become widely available from the software and sensor technology intertwined with the automation process. From a simple photoelectric sensor that detects whether an AS/RS shuttle has entered the proper level of the rack to a 2D LiDAR system looking down over a conveyor to understand the flow of parcels over time, real-time data is constantly surging into a PLC, IPC, or data collection tool within the warehouse. How this real-time data is used is one of the most interesting portions of the process.
As with any newer process or idea in a long-standing industry, challenges and speculation will arise. The transition from preventive to predictive maintenance is not free of these challenges. Where the data comes from, how to process it, and how much upfront implementation time it takes are all common questions during development.
In many cases, some of these questions can be answered using the data from the sensors already living in the warehouse. For example, it is becoming more difficult to walk into a warehouse without seeing the major use of automation, which requires sensors. To implement preventive maintenance, the data from an IO-Link capable sensor can be used as an analytical tool to understand when the machine could fail.
Now that the data is located and available, networking technology such as a managed Ethernet switch can transfer the data to the next step. A gateway can also be added if the data needs to be converted to a different programming language. The final step is employing a digital transformation strategy to examine the information in a centralized repository.
It is essential to remember that not every problem has a standardized solution, and a new solution can create a new silo of data. The strategy must ensure that every portion of the warehouse with data — from the AGV that carries a tote to the conveyor that transfers the tote to the production worker — behaves as one system.
After this convergence is coordinated, valuable preventive maintenance and condition monitoring analysis solutions can finally occur. A digital dashboard is a simple way to make informed decisions using the available data. To avoid making data analysis difficult, an open architecture solution can converge the data from the various systems to produce a much clearer view in the digital dashboard. When analytics highlight a problem, data sources can be analyzed to detect which system generates the issue, from micro-stops to AGVs receiving incorrect instructions.
In the end, a process analysis can be overlaid on top of the productivity metrics to get a picture of the overall health of the operation and a clear understanding of where the process can improve or where the next bottleneck or failure will occur.
A critical component of this data-driven, preventive maintenance solution is to share all the data sources from the smart sensors and machines to constantly evaluate the health and condition of the process. This is typically managed in the background of the operation in a non-time-sensitive fashion. The advantage is that rather than having the maintenance team periodically or regularly check the systems, the software-driven solution queries the smart sensors for their IO-Link-provided health diagnostics. This data can be shared to help operations, engineering, and management understand the warehouse’s long-term health. This requires the previously mentioned methods of converging systems and the technology connectivity that provides the data to be analyzed.
As we continue to debate which diagnostic approach is right for each specific warehouse organization, there are two aspects most may agree on today: having the ability and resources to prevent downtime and failure is strongly desired, and it’s a very interesting problem to solve.
An analogy for preventive and predictive maintenance
For most of my life, I pushed myself while working out at the local gym. I used my feelings to balance reasonable rest periods between exercises while keeping my heart rate elevated. After training sessions, I self-evaluated and asked myself if I had trained hard enough or could have pushed a little more. Then, I adjusted the next training session’s intensity, higher or lower, based on my feelings. Over time, I saw gains, which convinced me that my method was successful.
However, was the method optimal? Was there a more efficient path to success or progress? The answer is yes.
Now, instead of using my feelings, I use a smartwatch to record my heart rate and manage rest between exercises. Instead of guessing whether I’m having a rigorous training session, I check my calories burned and make informed decisions based on real-time data.
This example is how I like to think of preventive maintenance compared to predictive maintenance.
Preventive maintenance allows me to make decisions based on my feelings. Predictive maintenance allows me to make more accurate assessments based on real-time data and a simple dashboard.
You may also like:
Filed Under: Warehouse automation, Material handling • converting, Sensors (position + other), Sensors (proximity)