Robot Preventive Maintenance Planning: Calendar, Usage, and Condition-Based Service

When a robot arm stops mid-cycle on a Friday afternoon, the maintenance strategy you chose months ago suddenly matters more than any spec sheet. The question is not whether to do preventive maintenance. It is what should trigger it: the calendar, the hour meter, or the actual condition of the machine.

Most buyers pick one approach, apply it across the fleet, and move on. That works until it does not. A calendar-based schedule that over-serviced one robot may under-service another running double shifts in a humid environment. This article breaks down the three trigger strategies, what standards and OEM practices say about them, and what to verify with your supplier before committing to a plan.

Three Ways to Trigger Maintenance

Maintenance literature divides strategies into a spectrum from reactive to proactive. At one end is run-to-failure: fix it when it breaks. At the other is condition-based maintenance (CBM), where sensors and data tell you the machine needs attention before it fails. In between sit calendar-based and usage-based approaches, the two most common triggers in industrial robotics.

Calendar-Based: Simple but Blunt

Calendar-based maintenance triggers service at fixed time intervals: every 90 days, every six months, once a quarter. It is easy to plan, easy to budget, and easy to audit. If your operation runs a predictable single shift with consistent loads, a calendar schedule may be perfectly adequate.

The weakness is obvious. A robot running 16 hours a day in a welding cell accumulates wear far faster than one doing light pick-and-place for eight hours. Treating them the same means you either replace parts that still have life, or you miss degradation that has already started. Calendar-based plans also tend to be conservative, which means higher parts costs over time.

Usage-Based: Closer to Reality

Usage-based maintenance triggers service on operating hours, cycle counts, or parts processed. Instead of “every 90 days,” the trigger becomes “every 5,000 cycles” or “every 2,000 operating hours.” These are illustrative examples, not recommended intervals. This is closer to how wear actually accumulates.

The catch is data. You need reliable hour meters or cycle counters, and someone has to actually read them and act on the readings. In practice, usage-based plans work well when the robot controller logs operating hours automatically and the maintenance team has a process to review those logs on schedule. Without that discipline, a usage-based plan quietly degrades into a calendar-based one, or worse, into run-to-failure.

Condition-Based: Only When Needed

Condition-based maintenance uses periodic or continuous measurements, such as vibration, temperature, and motor current, to assess degradation and inform service decisions. With suitable thresholds and a response process, it can help detect problems early and avoid unnecessary component replacement. Monitoring does not guarantee that every failure will be predicted.

CBM requires appropriate measurements, data analysis, and a response process. It may be supplied by an OEM, an integrator, or a specialist monitoring provider. Compatibility, data access, and support must be confirmed for the exact robot and configuration.

What Standards Say

ISO 55001:2024 — Asset Management Framework

ISO 55001:2024 sets requirements for an asset management system. Its revisions address decision-making, risk and opportunity, data, knowledge, and life cycle management. The standard provides an organizational framework for managing assets rather than a robot-specific service schedule.

The new predictive-action section should not be read as an instruction to replace calendar maintenance with sensor-based maintenance. ISO technical committee guidance on the 2024 changes describes adapting to internal and external changes across risks, opportunities, services, and assets. Buyers still need maintenance instructions appropriate to their equipment and operating conditions.

ISO 17359:2018 — Condition Monitoring Guidelines

ISO 17359:2018 gives general procedures to consider when establishing a condition monitoring program and applies to all machines. It provides a framework for organizing monitoring, rather than a universal set of robot alarm thresholds or maintenance intervals.

For buyers, ISO 17359 is useful as a reference framework. It tells you what a structured condition monitoring program should look like, which helps when evaluating whether a supplier's CBM offering is comprehensive or superficial.

How OEMs Actually Do It

ABB Robotics: Data-Driven CBM

In its 2021 description of its Condition-Based Maintenance service, ABB explains how operational variables, including duty, speed, acceleration, and gearbox wear, are compared with its robot database to assess potential faults. The resulting robot-specific report includes analysis and maintenance recommendations. Ask ABB to confirm current availability and compatibility for your fleet.

A service description is not a reliability guarantee for your application. Request model-specific maintenance instructions and the operating conditions behind any reliability or service-interval claim.

FANUC: Zero Down Time (ZDT)

FANUC Zero Down Time (ZDT) collects robot health and operating data to support condition monitoring and maintenance alerts. Buyers should confirm controller compatibility, monitored components, connectivity requirements, and who responds to an alert before relying on it in production.

KUKA: Flexible Maintenance Contracts

KUKA service offerings include maintenance, spare parts, on-site support, and technical assistance. Confirm the local service scope, response times, and contract terms for your equipment. Offline simulation and cell planning tools should not be treated as evidence that a robot has condition monitoring.

Choosing a Strategy: What Buyers Should Verify

No single strategy is universally correct. The right approach depends on your application, duty cycle, environment, and what your supplier actually supports. Here is what to verify before committing:

What the supplier offers. Does the OEM provide CBM services? If so, what does it cover: vibration only, or also temperature, motor current, and gearbox condition? Is CBM included in the base package or sold as an add-on service? ABB's CBM and FANUC's ZDT are OEM-specific offerings, not industry-standard features that every robot ships with. For multi-site fleets, also consider how after-sales coordination aligns with the supplier's service model.

How maintenance intervals were determined. Ask what test conditions produced the recommended intervals. Were they tested under your load, environment, and shift pattern? A maintenance interval validated for single-shift, indoor, moderate-load operation may be inadequate for a two-shift welding cell in a humid facility.

What you can do yourself versus what requires authorized technicians. Some maintenance tasks are designed for user execution: visual inspections, filter changes, basic cleaning. Others may require authorized or factory-trained technicians, particularly anything involving safety functions, calibration, or internal mechanical components. Get this breakdown in writing before purchase, not after a warranty dispute.

Real replacement conditions for consumables. Lubricants, filters, tires, brushes, batteries: these have real replacement conditions that depend on usage and environment. Do not accept a fixed interval without understanding the conditions behind it. Ask for the data or at least the reasoning. For fleet-wide planning, see spare-parts planning for multi-site robot fleets.

Whether condition monitoring is available for your specific model. Not every robot in a supplier's lineup supports CBM. The flagship model may have it; the budget model may not. Confirm what is available for the exact configuration you are buying.

Common Pitfalls

Confusing maintenance labels. Scheduled maintenance, condition-based decisions, and failure prediction can overlap within one program. Ask what data is measured, whether the system assesses current condition or forecasts deterioration, and what action an alert requires. A dashboard alone does not establish predictive capability.

Assuming condition monitoring is included. Availability, compatible controllers, access to data, and fees depend on the offering and configuration. Confirm these details in the proposal rather than assuming every robot includes an OEM monitoring service.

Extrapolating from one application to another. Maintenance needs are application-dependent. A robot doing precision assembly in a clean room has different wear patterns than the same robot palletizing in a dusty warehouse. Do not assume the supplier's recommended interval applies universally across your applications.

A Practical Framework

Rather than choosing one strategy for the entire fleet, consider a layered approach:

1. Start with the manufacturer's baseline. Use the recommended calendar or usage-based intervals as the starting point. Any deviation from the manufacturer's schedule should be discussed and agreed with the supplier, as it may affect warranty coverage.
2. Add usage tracking where possible. If the controller logs operating hours and cycle counts, use them to inform maintenance planning. They are more accurate than the calendar alone.
3. Layer in condition monitoring where available. If the supplier offers CBM for your model, use it to catch issues early and to inform maintenance decisions. Any changes to maintenance intervals based on CBM data should be validated with the supplier.
4. Review and adjust. Maintenance plans are not set-and-forget. Review failure data, near-misses, and actual component condition at each service event. Adjust intervals based on what you learn. For systematic incident tracking, consider robot fleet incident management practices.

The goal is not to find the single correct maintenance strategy. It is to build a plan that matches how your robots actually run, backed by data you can trust and a supplier that supports the approach.

What to Ask Your Supplier

• What maintenance trigger strategies does this model support: calendar, usage, condition-based, or all three?
• What condition monitoring data is available, and is it included or sold separately?
• What were the test conditions behind the recommended maintenance intervals?
• Which maintenance tasks can our team perform, and which require authorized technicians?
• What are the real replacement conditions for lubricants, filters, batteries, and other consumables?
• How does the supplier's CBM offering compare to what ABB, FANUC, or KUKA provide?
• Can we access raw condition data, or only the supplier's interpreted alerts?