You are likely aware of the concept of condition based maintenance (CBM): essentially, performing the optimal amount of maintenance only when required (and on your own schedule). Much progress has been made on CBM lately, including many successful program launches. Yet, these programs are not without risk. When considering your CBM strategy, you should be aware of some of the potential pitfalls:
Not involving everyone
When starting a CBM system, a vital first step is to ensure all key stakeholders are involved, engaged and committed to the plan. This may seem trivial, but surprisingly, this is not always the case for CBM system deployments.
In one case, an engineer led a pilot program and installed sensors that fed predictive software, but the engineer only gained partial commitment from the maintenance staff. The program struggled and did not gain much traction. This could have been avoided had the maintenance staff been more involved on the front end.
Likewise, integration of the CBM program with the existing CMMS (or other specialized software) must be considered. The closer the two can mesh, the better. Your staff is probably comfortable using the CMMS to schedule and track maintenance. CBM will enhance this system, not necessarily replace it. Make the integration between the two as simple as possible, to ensure that your staff will use both systems effectively.
Scoping too large
Buzzwords like Smart Technology, IoT and Industry 4.0 may elicit a wide variety of reactions from industry professionals. These reactions can vary from skepticism to enthusiasm. However enthusiastic the team is, it is best to stay grounded in your approach. A CBM program is perhaps best viewed as a tool to help your operation. As with any tool, it can help make things easier, but it’s only valuable if used properly. Thus, a broad scope implemented rapidly can quickly become overwhelming and lead to a potential program failure. To fight this, start with a pilot to learn and adjust before going all in. You can steal a play from the agile methodology playbook to adapt small incremental steps quickly to avoid early setbacks.
Quantity over quality
An ambitious and cost-conscious engineer may be tempted to purchase as many low-cost sensors as possible and place them everywhere in the plant. This potentially introduces a new problem of sensor administration, where more time and effort are spent in replacing cheap sensors than in using them for improving maintenance. Consider the recent discussion in “Do multiple sensors improve accuracy?” Rather than multiplying sensors to gather extra data, be more strategic in your approach. Use quality sensors to gather the most relevant information first. Further sensors can be added as the needs are identified.
No sensor management strategy
With the potential influx of new sensors, you must consider your sensor management strategy. For example, how often will you be checking or calibrating them? Will they feed data through PLCs, DCS, or separate network? You will want to find a good balance between adding burdensome PM procedures, yet gaining value from your new data insights.
The use of wireless sensors adds in additional complexities. A battery strategy will need to be thought through. Consider how you will check your device battery life, and at what threshold you will respond. Who will be responsible for this task? A good CMMS integrated with your CBM sensors will help solve a lot of these complexities. All these questions should be answered as part of your CBM deployment.
Security strategy not considered
The proliferation of wireless technology has allowed for cost-effective and rapid deployment of CBM. Particularly in areas with electrical classification requirements, the cost savings in avoiding specialized wiring are considerable. Within the past five years, the possibilities of electronics within manufacturing have increased substantially.
Even with stringent requirements of Safety Instrumented Systems (SIS), there is possibility of using wireless technology. However, there is great risk in allowing communication between PLCs and business LANs. The possibility of manipulating a machine from an external network must be eliminated. A good understanding of proper network architecture (DMZs, process LAN versus business LAN, etc.) should be a requirement of your CBM strategy.
Unfortunately, this is a subject too large to fully cover in this article. A reference such as this white paper can help you get started.
The explosion of IoT related technology has made a tangible condition based monitoring program more possible than ever before. However, one cannot simply install equipment and software and expect immediate results. As with any new endeavor, there are difficulties that one can encounter during development and deployment. These difficulties can be related to hardware, network or culture. To truly realize all benefits, an overall strategy for CBM must be carefully considered.