Community benefits from sensing systems such as motorist way finding, demand based pricing, directed enforcement, dynamic pricing and meter reset are all dependent on accuracy. How Fybr measures accuracy goes far and beyond current standards, though. We believe to provide value, performance measurements for sensing systems should:
• Represent real world uses of the system
• Apply to any system for comparison purposes
• Be measured over time to ensure continued compliance
• Provide a clear and understandable process and result
Understanding performance is about much more than a single factor, however. At Fybr, we believe four key metrics are necessary to properly evaluate a sensing system’s performance. Each of these metrics taken separately tell a part of the story, but in concert, they give the evaluator the best view of real world system performance.
1. Event Capture
Simply Defined: Event Capture is used to ensure the sensing system is capturing all real parking events and reporting them correctly. Systems that consistently perform above 95 percent are capable of producing meaningful and valuable data that can be used for setting dynamic parking pricing and other community benefits.
How It Works: Parking events for a time period are captured either by a live person or by using a time stamped video camera. The actual parking events are then compared to the parking events registered in the system for the same time period.
2. Elapsed Occupancy
Simply Defined: Elapsed Occupancy considers if the sensing system correctly represents space occupancy. Elapsed occupancy is especially important for understanding the validity of way finding and when the system is being used to make policy decisions based on occupancy rates.
How It Works: Elapsed occupancy is measured by calculating a percentage of the total test minutes that were correctly represented.
3. System Latency
Simply Defined: System latency demonstrates that the underlying system (sensors, network, back end processing) is operating consistently and correctly.
How it Works: Latency is calculated by looking at the difference between timestamp from the sensor’s real time clock and the event being received, evaluated and recorded in the system’s database.
4. System Uptime
Simply Defined: System Uptime is a quantitative measurement that proves the reliability of the of the parking system’s underlying technology. It provides the final piece of the puzzle when looking at the comprehensive performance of a sensor system.
How it Works: Uptime is calculated by looking at the number of minutes sensing services were available, divided by the total number of minutes in the month and excluding any scheduled downtime.
For a fuller understanding of the relative performance of parking sensing systems, read our white paper, Measuring Parking Sensor Performance, authored by Fybr Chief Operating Officer Thomas Valley.