Data-Driven Maintenance (DDM)

Imagine being able to evaluate every maintenance work order executed, every corrective notification, and every condition diagnostic from CBM (Condition-Based Monitoring) and automatically orchestrate what needs to be done, how, and when. Only do the right things, in the right way, at the right time—continuously learning and optimizing maintenance from within.

This article explores how new BOAT technology can achieve this by applying business logic and utilizing existing data, systems, and organizational resources.

"There is nothing so useless as doing efficiently (or improving) that which should not be done at all” Peter Drucker.

BOAT (Business Orchestration and Automation) is Gartner's brand-new software category. It enables enterprises to layer business logic on top and autonomously orchestrate and automate parts of or entire end-to-end business processes. AI engines are embedded to assist in continuous learning and improvement.

PDCA (Plan-Do-Check-Act) is an improvement cycle based on the scientific method of proposing a change in a process, implementing the change, measuring the results, and taking appropriate action. It is also known as the Deming Cycle after W. Edwards Deming. PDCA is the foundation of continuous improvement or kaizen.

Orchestr8 is a new, modern, and genuinely ground-up BOAT technology developed by Insert:Logic. It is a robust general-purpose automation platform built around autonomous orchestration, integration, and AI. Built on open standards, it ensures minimal lock-in as you layer things on top, building and scaling solutions leveraging in-house competence, systems, and data.        

Right maintenance is essential for asset-intensive industries' availability, safety, and cost. Let's begin with some common maintenance challenges and how BOAT technology can help address them:

From:

• Planned maintenance occurs on "autopilot" year after year (same interval, method, checklist, time estimate...).
• The operational unit feels that we continue to do much of what we have always done, repeating the same tasks (even though they do not create any value), making the same mistakes, not correcting inaccurate work orders, etc., all of which make the workday ineffective and inefficient.
• Corrective notifications often come in long text formats, requiring back-and-forth dialogue to confirm intent. A large portion goes to planning without being critically assessed.
• Investments in predictive and prescriptive condition-based monitoring result in a number of dashboards and information someone has to interpret.
• Suggestions for improvements tend to accumulate (and not be executed).
• The volume of planned and corrective activities and all the changes make planning and replanning work complex and burdensome.
• Learning and correction require significant effort and occur, at best, ad hoc.

To:

• All work orders are automatically evaluated upon completion. Potential improvements are identified and sent to the appropriate role for approval. Upon acceptance, the underlying systems are automatically updated.
• Each corrective notification is captured in an intelligent format and then routed to the correct role for evaluation. It is only accepted if it is deemed reasonable. If accepted, it is combined with other work whenever possible.
• Suggested improvements are promptly assessed and implemented if approved.
• We learn from every task, correcting it to ensure it is correct the next time and eliminating anything that doesn't make sense.

How does it work, and where do you start?

Why is high precision continuous learning and improvement so essential? Things get right when we continuously learn and improve based on everything we do. We obtain the correct data, perform the right activities, and establish the right intervals and work packages (hours, equipment, parts, work procedures, etc.). Most importantly, we eliminate everything that represents no value. This, in turn, has a chain effect on planning/scheduling, our ability to concentrate on the right tasks, and our ability to have a more efficient working day in the field.

Let's follow the PDCA cycle, starting with the Check step.

Check:

The BOAT solution monitors all completed work orders in your ERP, maintenance system, and/or data platform. It checks for any issues or deviations from the planned work order and if it is safety or production-critical. If not safety or production critical, and no issues are found, it checks if any problems were detected in previous intervals (for example, the last ten intervals). If no issues are found, BOAT sends a proposal and relevant information to the appropriate role or function for changing the interval or deactivating it. The exact same process is followed when deviations from the plan are detected, such as time spent, tools, materials, method, etc. It also coordinates the proper workflow if multiple roles or functions need to be involved.

For corrective notifications, BOAT assists the maintenance worker in determining whether and how to notify. If notification is necessary, BOAT evaluates the information before sending it to the responsible role for approval.

When integrated with CBM (Condition-Based Monitoring) systems, BOAT dynamically orchestrates what to do, when to do it, and who to involve based on condition diagnostics.

Now, the Chech loop is spinning. With the help of BOAT, we can filter out and check everything relevant. Because it is intelligently orchestrated—only relevant roles are involved in each case—making it easy to manage as part of the daily job.

Act:

When the automatically generated proposals to adjust or deactivate planned maintenance activities, create corrective notifications, or improve the way of working reach the proper role/function, the role/function evaluates whether it should be accepted, changed, or rejected. Depending on the outcome, BOAT updates the relevant underlying systems (ERP/Maintenance/etc.). The Act step can be done entirely automatically if human control is unnecessary.

Now, the Act loop is spinning. We learn and adjust continuously, our maintenance system is increasingly accurate, and we only do the right things in the right way and at the right time.

Plan:

Gradually, as BOAT assists the organization in weeding out everything that does not make sense and only doing the right things in the right way and at the right time, the planning complexity gets significantly reduced. Time can be spent on the right things that create value.

Since the work package specifications (time and cost estimates, required materials and tools, safety instructions, step-by-step procedures, etc.) are becoming increasingly correct, the degree of automatic work preparation and scheduling is increasing.

The underlying AI engine supports dynamic bundling, scheduling, and planning.

Now, the Planning loop is spinning. The combination of fewer plan components, automatic work preparations, automatic scheduling, and bundling of work orders (dynamically AI-assisted as changes occur) makes a huge difference.

Do:

Finally, the maintenance workers can sense the shift—how work is properly prepared and planned, how the working day becomes much more efficient and meaningful, and how they are involved in improving their way of working and making a difference.

Now, something is happening. It's easy for everyone to engage and make suggestions for improvements. Within hours/days, it′s evaluated and implemented. Your contributions feel valued. When this spreads throughout an operational organization, a collective hands-on improvement power of great value is mobilized.

Finally, it is worth mentioning how AI can now play an increasing role. With a data-driven way of maintaining/operating through BOAT, the AI engine gathers data from each step in the process for each case —every finding, activity, routing, involvement, and decision—and continuously feeds suggestions for improvement based on this data.

The four improvement loops in the PDCA cycle and the cycle as a whole are now spinning fast.

BOAT is here, and it matters.

Read more about BOAT: https://www.insertlogic.io/boat

#BOAT #businessorchestration #automation #AI