Positioning Accuracy
The degree of correctness in position measurements, encompassing both systematic and random errors. Should specify measurement conditions and confidence levels. Accuracy varies across coverage area, typically best near anchors and degrading at edges.
Measure of how precisely RTLS determines object positions relative to true positions, fundamental performance specification. Accuracy expressed as: mean error (average distance between reported and true positions), percentile error (eg., 90th percentile: 90% of measurements within specified error), or error distribution showing error characteristics. Critical to specify conditions: static versus dynamic (moving tags typically show 20-50% worse accuracy), line-of-sight versus non-line-of-sight (NLoS degrades accuracy 30-100%+), and confidence level (50th percentile/median, 90th percentile, 95th percentile - vendor specifications often cite median which is optimistic compared to 90th percentile). Technology accuracy ranges: (1) UWB - 10-30 cm (90th percentile, good conditions), 20-50 cm (industrial environments with some NLoS). (2) BLE AoA - 0.5-2 meters (good conditions), 1-3 meters (realistic industrial). (3) BLE RSSI - 3-8 meters (highly variable, environment-dependent). (4) Wi-Fi fingerprinting - 3-15 meters (depends on fingerprint database quality). (5) Active RFID - zone-level only (5-20 meter zones). (6) GPS - 3-10 meters (outdoors, civilian signals). Accuracy requirements vary by application: (1) Centimeter-level (5-30 cm) - AGV guidance, robotic picking, precision assembly tracking. (2) Decimeter-level (30-100 cm) - forklift tracking, collision avoidance, work cell tracking. (3) Meter-level (1-5 meters) - warehouse zone tracking, personnel location, large area asset tracking. (4) Room/zone-level (5-20 meters) - basic asset location, presence detection, occupancy counting. Example: vendor claiming 15 cm accuracy might mean median in ideal LoS conditions, while 90th percentile in actual environment might be 40-60 cm. Understanding accuracy critical for setting requirements, evaluating vendors, managing expectations, and troubleshooting performance issues.