SCADA Comms SLA Reporting: What to Measure for a Radio Network, and How to Report It to Ops Leadership
A radio network that nobody measures gets managed by anecdote, and leadership remembers the worst night, not the trend. A small set of well-defined metrics, reported consistently, lets you show whether the network is getting better or quietly getting worse.
Decide what "up" means first
Before any number is useful, define the terms. Two teams measuring "availability" can get very different answers if one counts a link as up whenever the radio is powered and the other counts it up only when SCADA data is flowing.
Settle these definitions in writing:
- What is the unit? A radio link, a remote site, or a data point. Site-level (can SCADA see the site) usually matters most to operations, while link-level helps the comms team diagnose.
- What counts as down? For example, a defined number of consecutive missed polls, or no valid data for a set period. Choose a rule and keep it stable so trends stay comparable.
- What is excluded? Planned maintenance windows and customer-side outages are often excluded from an SLA figure, but be explicit and report the excluded time separately.
- What is the reporting period? Monthly and quarterly are common; trailing 12-month views show the long trend.
The four core metrics
1. Availability
The percentage of the period a site or link was up under your definition. Small differences in percentage are large in time. Over a year, 99.5 percent availability allows about 43.8 hours of downtime, 99.9 percent allows about 8.8 hours, and 99.99 percent allows only about 53 minutes. Always show hours or minutes alongside the percentage, because leadership understands the former better. Report the worst sites as well as the average: a network average of 99.9 percent can hide one site that is down for days.
2. Link margin trend
Availability is a rear-view measure; margin trend is the more forward-looking one. Track fade margin (received signal level minus receiver threshold) per link over time, and report how many links have margin below a threshold you choose, along with how that count is changing. A healthy availability figure with a falling margin trend is a warning that today's good number will not hold. This is the metric that moves reporting from "what broke" to "what will need attention."
3. MTBF (mean time between failures)
The average operating time between failures, computed per link or per equipment class. It helps identify chronic problem sites and equipment types, and informs spares and replacement planning. Keep in mind the caveats: with few failures the average is statistically shaky, and a mean can mask a single very bad site. Pair it with mean time to restore, the average time to bring a failed link back, measured from detection to restoration, because MTBF and the mean down time together set availability (availability = MTBF / (MTBF + mean down time)). Strictly, mean time to repair counts only active repair; for an availability figure you want the fuller down time, and the sooner you detect a failure the smaller that down time is.
4. Comms-loss events
A count of discrete outage events, with duration, site, and cause category (radio hardware, power, path or weather, interference, third-party, unknown). Counts and durations tell different stories: ten brief dropouts and one long outage can give the same downtime but need different fixes. The cause category is the most valuable column, because it separates fixable patterns from bad luck. Be honest about "unknown", and work to shrink it over time.
Useful extras
- Mean time to detect. How long a failure existed before someone knew about it. Long detection times are a monitoring problem, not a radio problem.
- Planned versus emergency visits. The ratio of scheduled to unscheduled field trips is a good indicator of how proactive the program is, and ties comms performance to a cost leadership understands.
- Repeat offenders. The top five or ten sites by event count or downtime, with the action planned for each.
- Dependency exposure. Which repeaters or hubs, if lost, take the most sites with them.
How to report it to ops leadership
- Lead with a one-page summary. Availability against target, the trend (up, flat, or down), the number of comms-loss events, and the three things that need a decision or funding.
- Use plain language and consistent visuals. One chart of availability over 12 months and one of margin trend say more than a table of figures. Use the same format every period so changes stand out.
- Connect to operations. Translate downtime into what it meant: hours of manual readings, number of dispatches, any production or measurement impact. Leaders respond to operational consequences.
- Show cause, not just count. A short list of root-cause categories shows where improvement dollars go.
- Separate what happened from what is coming. One section for the past period, one for links with shrinking margin and the planned visits to address them.
- Be honest about the data. Note gaps, sites not monitored, and definition changes. Trust in the report is worth more than a flattering number.
- Set targets that match consequence. A custody-transfer site and a minor monitoring point should not have the same target. Tiered targets are easier to defend and to meet.
Add margin trend to your next SLA report.
Aevus Comms (RF Link Intelligence) is designed to supply the raw ingredients for the forward-looking parts of that report: cross-vendor fade-margin trends and a record of link-degradation and comms-loss events, in one read-only view that can feed the reporting you already do.
Related Articles
The leading indicator behind the margin-trend metric.
Bands, degradation modes, and redundancy for SCADA radio.
Ring topologies, PRP, HSR, and the failover math behind uptime.
Ready to see how this applies to your operation? Start a pilot conversation, no commitment, no field changes.