Definition
Average gap between SLA-reported response timestamps and the timestamps when meaningful diagnostic or repair work actually commenced. Monitoring precision creates an incentive for premature status changes that satisfy the contract while the work starts later.
The Measurement Gap
The Measurement Gap
O&M service contracts typically include response time SLAs: the provider must have a technician on site within 4 hours for Priority 1 events, 24 hours for Priority 2, and so on. AI monitoring systems track compliance by recording when the CMMS work order status changes from "dispatched" to "on site" or "work started." The provider is measured against these timestamps. The gap is that the timestamp records a manual status change, not a verifiable activity. A technician can mark "on site" from the parking lot, or even from a nearby location, at 3 hours and 55 minutes to meet the 4-hour SLA, then spend another 90 minutes on travel, preparation, and tower access before meaningful work begins.
Under manual contract monitoring, this gaming was difficult to detect because no one compared status change timestamps against actual work activity timestamps. Under AI monitoring, the SLA compliance is tracked to the minute, which creates both the incentive (precision monitoring means every minute matters for compliance scoring) and the opportunity (a single timestamp click satisfies the monitoring system). The actual mean time to repair is longer than the reported MTTR because the work start clock runs from the status change, not from when tools actually touched the asset.
The Decision It Enables
RTAS enables the asset owner to distinguish between authentic response (the provider genuinely responds within the SLA window and begins work immediately) and timestamp-managed response (the provider satisfies the monitoring system's timestamp check while actual work begins significantly later). When RTAS is near zero, the SLA timestamps are authentic and the reported response time reflects reality. When RTAS is consistently above 30 minutes for priority work orders, the SLA compliance metric is measuring the provider's timestamp management rather than their operational responsiveness. This finding drives two potential actions: renegotiating the SLA measurement methodology to require verifiable activity triggers (SCADA login, parts issuance, diagnostic reading) rather than manual status changes, or adding an RTAS-based quality adjustment to the SLA compliance calculation.
Formula
Primary Formula:
RTAS = (1/N) x SUM(Actual_Work_Start_i minus Status_Change_i) measured in minutes
Variable Definitions:
| Variable | Definition |
|---|---|
| Actual_Work_Start_i | the timestamp of the first verifiable work activity for work order i. Verifiable activities include SCADA system access or login, tool checkout or equipment connection, parts pull or issuance from ... |
| Status_Change_i | the timestamp when the CMMS work order status changed to "on site" or "work started" for work order i, which is the timestamp used for SLA compliance measurement |
| N | the number of work orders in the measurement sample, which should focus on SLA-penalized priority levels (typically P1 and P2 events) |
Calculation Example:
A sample of 40 P1 and P2 work orders is analyzed. For each, the CMMS status change timestamp is compared against the earliest verifiable activity timestamp from SCADA access logs, parts issuance records, or mobile application activity. The average gap is 42 minutes. RTAS = 42 minutes. For 12 of the 40 work orders, the gap exceeds 60 minutes. For 6 work orders, the status change occurred more than 90 minutes before any verifiable activity, suggesting the technician marked "on site" well before arriving at the asset.
Edge Cases and Measurement Boundaries:
Travel within the site from entrance gate to the specific asset is a legitimate delay that contributes to RTAS without indicating gaming. Define "on site" at the appropriate granularity for the asset type. Waiting for weather windows after arrival is a legitimate delay; flag these separately as "on-site weather wait" and exclude from the gaming analysis. Night arrivals for morning work (technician arrives at lodging near the site the evening before) should be excluded from RTAS because the delay is planned and legitimate. Track RTAS specifically for work orders where the SLA window was nearly exhausted (response at 3:50 for a 4-hour SLA) because these represent the highest gaming incentive.
Proposed formula, pending validation pilots.