Historical Data

Almost every building system produces a record of the past. The most useful types for facility teams are:

• Sensor and BAS trend logs: temperatures, pressures, valve positions, fan speeds and on and off status
• Energy and utility data: meter readings, interval data and monthly bills
• Maintenance records: past work orders, repairs, parts used and labor hours
• Alarm and event history: when alarms occurred, how long they lasted and how they were resolved
• Inspection and test records: fire protection, elevator, generator and backflow tests
• Occupancy and space use records
• Capital and replacement history: when equipment was installed, rebuilt or replaced

Most building historical data starts as a trend log in the building automation system, where a controller records a point's value at a fixed interval, such as every 5 or 15 minutes, or whenever the value changes by a set amount (change of value, or COV). Many controllers keep a limited number of samples on board, so older data is overwritten unless it is archived to a server, a historian database or a cloud platform.

Maintenance history lives in the CMMS, energy data in the utility or metering system, and documents in shared drives or document management systems. Because each system stores its own history in its own format, lining them up over the same time period is often the hardest part of any analysis.

Historical data turns a single reading into context: whether a value is normal, new, seasonal or getting worse.

• Troubleshooting: comparing current behavior with how a system ran before a problem started
• Fault detection: finding equipment that runs when the building is unoccupied, simultaneous heating and cooling, or leaking valves
• Energy baselines: setting normal consumption so the result of an upgrade can be measured, as described in ASHRAE Guideline 14
• Maintenance planning: spotting assets with repeat failures and adjusting preventive maintenance intervals
• Capital planning: using age, condition and repair costs to decide when to replace equipment
• Compliance and reporting: showing that inspections, tests and setpoints met requirements

Historical data is useful when it can be trusted. Common problems include gaps when a controller or network was offline, sensors that drifted out of calibration, points that were renamed or reassigned during a controls upgrade, inconsistent units, and timestamps that shift with time zones or daylight saving time. Maintenance records have their own issues, such as work orders closed without notes or logged against the wrong asset.

It helps to agree on point naming, keep trend intervals consistent for key points, and record changes to equipment and controls so later analysis can account for them.

Real-time data shows what is happening now, such as the current supply air temperature. Historical data is the record of those readings over time. Real-time data drives alarms and day-to-day responses, while historical data explains patterns and supports longer-term decisions.

Syyclops connects to the building automation system through a read-only BACnet gateway that creates trend data, and brings in work orders and maintenance history from the existing CMMS. Its AI Engineer uses that historical data, together with asset conditions and past maintenance records, to help teams troubleshoot.