Cutting a maintenance budget is not the same as reducing maintenance cost. Deferring an inspection may improve this month’s budget while increasing next quarter’s downtime. Defensible optimization removes work that does not control a meaningful risk, improves work that does, and then checks the consequences for equipment. Start with a bounded scope: a pump family, one line, or the assets responsible for the largest losses.
Establish a baseline before changing the plan
Collect preventive and corrective hours, parts, contractors, downtime and output over a representative period. A full year often helps reveal seasonality, but a recent process change can make that year a poor comparator. Normalize results by operating hours or output. Separate accounting expenditure from released capacity: recovered time does not automatically become a cash saving.
Connect each task to a failure mode
For every task, ask which function it protects, which failure it addresses, and how it detects or prevents that failure. “Check motor” is not specific enough: a check needs a target condition, method, threshold and expected action. Duplicate tasks, inspections without acceptance criteria and periodic replacements without justification deserve review; that does not mean they can be deleted immediately.
- Does the task address a legal obligation, protective function or manufacturer requirement? Validate those constraints before any change.
- Can the targeted defect be detected early enough to act? Compare detection lead time, parts availability and preparation time.
- Does the interval fit the degradation mechanism and actual operating conditions?
- Does the inspection result trigger a documented decision, or is it simply filed away?
- Is the task duplicated elsewhere, and could the intervention itself introduce a defect?
A worked calculation, not a promised result
Fictional teaching example: a plan uses 1,200 hours annually. A technical review proposes removing 240 hours of duplicate tasks and adding 60 hours of targeted monitoring. The balance is 180 hours released. At an assumed labour cost of CAD 90 per hour, the capacity value is CAD 16,200 annually. If staffing remains unchanged, this is not an automatic payroll reduction; it may help clear corrective work. Calculate parts savings separately.
Pilot the change and watch for unintended effects
Have technical, maintenance and operations owners approve the change. Retain the old plan and the rationale for every decision. During the pilot, track detected defects, failures per operating hour, urgent work and availability. Set rollback criteria before starting. Lower preventive hours accompanied by more emergency repairs are not a success.
What Vizory’s published projects show
Vizory’s published Rio Tinto Fer & Titane defect-elimination case reports a 65% reduction in maintenance-plan costs. That result belongs to that specific engagement; it is neither an average nor a target for every plant. The case page provides the published context. For your site, ask for a measured baseline and a change register rather than a percentage announced before analysis.