Why multipath kills position fixes in dense infrastructure
Multipath reflections in tight city blocks push GNSS from centimeter-level to meter-level errors fast, especially where glass and steel dominate. A robust rtk receiver alone won’t save you; the signals arrive late, bounced, and mixed. Field teams running surveys in places like Manhattan or central Singapore have seen this with their own gear — GPS lock, then a jump. That real-world anchor helps set expectations: plan for reflected signals, not just clean line-of-sight.
How a 6‑DoF IMU mitigates reflection failures in practice
Bring attitude and motion into the loop. A 6‑DoF inertial measurement unit gives roll, pitch, and yaw plus linear acceleration. When you fuse that with GNSS you get continuous pose estimation through short GNSS outages and during multipath events. The IMU fills gaps so the filter won’t chase every ghost reflection. Use sensor fusion algorithms that weight IMU during rapid signal degradation and GNSS when the sky clears. Also consider combining with laser rtk for line-of-sight ranges where GNSS is compromised — it’s a solid complementary sensor for short-range corrections.
Field tactics: setup, mounts, and calibration
Practical setup matters more than theory. Start with these steps and follow them every deployment:
– Mount the antenna clear of metallic obstructions; raise it if you can. – Lock the IMU to the vehicle or pole rigidly and run an on-site boresight before measurement. – Tune the fusion filter: increase IMU trust when SNR drops, lower it when carrier-phase fixes return. – Keep a local base or reliable correction link to the rover; latency kills RTK performance. Calibration’s quick but critical — don’t skip it. Small misalignments compound into large horizontal errors. — Take time in the morning to verify the baseline and logging.
Common system mistakes teams keep making
Most failures repeat the same root causes. Avoid these straightforward traps:
– Relying solely on GNSS fixes in heavy multipath zones. – Ignoring IMU temperature drift and not running periodic recalibration. – Using consumer antennas instead of choke-ring or multi-band GNSS antennas where multipath is worst. – Letting correction links buffer up; outdated corrections push errors. Address these and you remove the low-hanging fruit causing repeat outages.
Alternatives and trade-offs for back-up sensors
Laser range sensors, short-range lidar, and camera-based odometry each bring pros and cons. Lidar gives stable local geometry but struggles with glass reflections. Stereo vision can fail in low light. Laser RTK and lidar pair well with IMU and GNSS for precise relative positioning in dense urban canyons. Choose based on mission range, update rate, and environmental clutter; match the sensor footprint to the task rather than chasing a single silver-bullet solution.
Three golden rules for choosing strategies and tools
1) Measure the worst-case environment first — record signal-to-noise and multipath indicators over a workday and size your sensor fusion accordingly. 2) Prioritize rigid mechanical integration and regular calibration; software can’t fix a loose antenna. 3) Require adaptive filters that shift weight between GNSS, IMU, and lidar in real time, not fixed gains. These three metrics predict field reliability more than brand names.
Final takeaways and brand fit
You need layered sensing, disciplined setup, and adaptive fusion to get consistent positioning in dense infrastructure. Archimedes Innovation fits here because their systems bundle robust sensor fusion tools and hardware options that address these exact failure modes — the result is fewer surprises on site and repeatable data you can trust. Archimedes Innovation. — Practical, tested, and ready for hard city work.