Domain 1 Article Series · Part 4 of 6 | RSF Whitepaper v1.0 · §3.4
Two identical robots, bought the same month, running the same annual maintenance contract. One works a single day shift in a clean assembly cell. The other runs three shifts in a foundry, next to the die-cast machine. Same calendar, same PM plan: one service visit every twelve months.
By its first anniversary, the day-shift robot has logged roughly 2,000 operating hours. The foundry robot has logged over 6,000 — in heat, with metal dust — and its reducers have consumed three times the wear life between the same two service visits. On paper the maintenance program treats these machines as equals. Physically, one of them is aging at triple speed, and the maintenance plan hasn't noticed.
This is the problem Phase 3 exists to solve, and it is why the RSF Whitepaper is pointed about what preventive maintenance actually is: not an insurance ritual performed on a schedule, but the mechanism that converts random fault risk into planned activity. The design question at the center of Phase 3 is deceptively simple — when should maintenance happen? — and most PM programs get it wrong before the first work order is written.
Three Tiers, Three Different Jobs
The RSF curriculum structures preventive maintenance as three tiers, distinguished not by thoroughness but by what each tier is for — and who is qualified to perform it:
| Tier | Trigger & Personnel | Core Content | What It Protects |
| L1 Weekly Inspection | Every 5–7 operating days · resident engineer or on-site RSF Professional | Visual inspection of cables, connectors, joints; lubrication check on high-cycle axes; alarm-log review; cooling system check; brief all-axes functional test; PM log entry | Early detection — catching the loose connector before it becomes an intermittent fault |
| L2 Scheduled PM | Per OEM operating-hour threshold, typically 1,000–2,500 hrs · RSF Professional or above | Torque calibration against IPR baseline; encoder accuracy verification; reducer oil change and gear-mesh inspection; cabinet internal and thermal check; firmware update review | Wear management — keeping measured drift within specification before it compounds |
| L3 Major Overhaul | Per OEM major interval, typically 20,000–40,000 hrs or 5–7 years · RSF Expert + OEM coordination | Full reducer replacement or rebuild; complete rewiring; control-board component-level work; full re-calibration against a new baseline; SLA renegotiation if the performance envelope changes | Life extension — resetting the machine's technical state rather than patching it |
Two details in this structure repay attention. First, L2's torque calibration is performed against the IPR baseline — the document from Part 2 of this series doing exactly the job it was created for. Second, L3 ends with a possible SLA renegotiation: RSF treats a major overhaul as an event that can change what the machine is contractually capable of, not just a big repair. The tiers are connected to the rest of the framework, not floating on their own.
Calendar, Runtime, Condition: Choosing the Trigger
The tier structure answers what gets done and by whom. The harder question is when — and RSF presents three strategy models, each with an honest cost:
| Model | Trigger | Best For | The Catch |
| Calendar-Based | Fixed time interval — e.g., every 12 months | Stable utilization, single-shift operation; simple to plan and budget | The interval ignores actual load. A three-shift robot on a calendar plan is under-maintained; a lightly used one is over-maintained |
| Runtime-Based | Cumulative operating hours reach an OEM threshold | All operating modes — the default model; tracks real wear far better than calendar time | Needs a reliable hour-meter — built into most modern controllers, but the data has to be read and acted on |
| Condition-Based (CBM) | Sensor data or a diagnostic metric crosses a threshold | Critical, high-OEE equipment; maintenance happens only when the data says so — least waste | Requires remote-monitoring infrastructure and higher upfront investment |
The foundry robot from the opening is a calendar-model casualty. Move it to a runtime trigger and the math changes immediately: at 6,000 hours per year against a 2,000-hour L2 threshold, it gets three scheduled services annually instead of one — not because anyone decided to triple the budget, but because the trigger finally measures the thing that actually wears the machine.
The models are not mutually exclusive, and mature programs usually layer them: runtime-based L2 as the backbone, calendar-based L1 for the weekly rhythm, CBM on the handful of machines where an unplanned stop costs more per hour than the monitoring system costs per year.
The Second Correction: Where the Robot Lives
Operating hours capture how much the robot works. They do not capture where. The whitepaper's PM design guidance adds an environmental correction on top of the strategy model, because the same 2,000 hours consume different fractions of component life in different conditions.
The pattern is consistent across environments: high ambient temperature accelerates lubricant degradation and stresses cabinet electronics — near heat sources, oil-change and filter intervals shorten. Conductive or abrasive dust — foundries, grinding cells — attacks seals and contaminates connectors, pulling inspection intervals in. High-humidity or washdown environments raise corrosion risk on connectors and grounding points. The specific multipliers are application engineering, set per site; the principle is fixed: an OEM interval assumes a reference environment, and a real deployment rarely matches it. The foundry robot needed both corrections at once — a runtime trigger for its utilization and shortened intervals for its atmosphere — and its calendar-based plan delivered neither.
What Each PM Leaves Behind
A PM visit that only replaces oil has done half its job. In the RSF structure, scheduled maintenance is also a measurement event: each annual PM closes with a configuration baseline snapshot — the PM record plus re-measured accuracy — filed as a time-series point in the equipment passport.
Over years, these snapshots become the machine's trend line. The commissioning baseline says what the robot was; the chain of PM snapshots says how it has moved. When a decision point arrives — overhaul or retire, the Phase 5 question — that trend line is the evidence the decision stands on. A PM program that fixes things but records nothing is spending money on maintenance while starving the decisions that depend on maintenance data.
Two Takeaways
For engineers preparing for RSF Professional certification: the L1 checklist and its cadence are execution-level material, and the three strategy models — including the failure mode of each — are core testable content in §3.4. Know why runtime-based is the default and what CBM requires before it can be promised to a customer.
For operations managers: pull one number before your next contract renewal — actual annual operating hours per robot, from the controller. Divide by the OEM's L2 interval. If the answer is 2.5 services per year and your contract says one, the gap between those numbers is wear nobody is managing. That is the three-shift robot on the one-shift plan, and it is more common than anyone in the room will want to admit.
Part 5 enters the phase everyone knows best and understands least: corrective maintenance — the eight-step process, the RCA trigger conditions, and the closed loop that separates organizations that fix robots from organizations that improve them.
Read the full Phase 3 specification RSF Whitepaper v1.0, §3.4 — PM strategy design, wear-parts management, and the maintenance documentation system. Free download. rsf.robottoday.com |
RSF (Robotics Service Framework) is an initiative of RobotToday.com. International standards cited are referenced for descriptive purposes; the official published text of each standards organization remains authoritative. Tier definitions and strategy models follow the RSF Professional training curriculum (Module 1), which operationalizes RSF Whitepaper v1.0 §3.4; operating-hour thresholds shown are typical ranges — OEM-specified intervals for each robot model remain authoritative. The opening scenario is illustrative. This article does not constitute a compliance certification document.
Leave a comment