Hickam AFB GPS Accuracy Testing
While were out at Hickam AFB for the NSIN - National Security Innovation Network Propel event we had a chance to run an accuracy test outside the hangar. We'd noticed that GPS on our mobile apps was pretty noisy on island and we were curious to quantify it with a test. We set up five static receivers and one rover. I could only fit one RTK unit in luggage ?? . Below are the improvements we saw in the Zephr solution compared to GPS in meters.
Across the entire test we were able to reduce the error from GPS by -7.4478 meters on average. Next we can look specifically at the roving receiver:
领英推荐
The blue line = Zephr geodetic error, orange line = Zephr total error, red line = GPS geodetic error and green line = GPS total error. For the rover we are improving to geodetic error from 3.18m to 54cm and the total error from 6.3m to 2.74m. Next we can look at one of the static receivers to see the specifics of its improvement.
Again the blue line = Zephr geodetic error, orange line = Zephr total error, red line = GPS geodetic error and green line = GPS total error. For the static receiver we improve total error from 12.46m to 1.88m and geodetic error from 9.2m to 64cm. It was interesting to see our qualitative experience of degraded GPS performance on our mobile apps manifested in the accuracy test. Also, it was really encouraging to see that the Zephr ensemble model could compensate for those degradations. My hunch is that degradation comes from increased atmospheric interference in Hawaii with it being closer to the equator, which the native Android Klobuchar model doesn't handle well, but that is purely a hypothesis. Stay tuned for some fun use cases of better accuracy soon!
Former Executive Secretary of the National Space Council Users' Advisory Group (UAG), NASA Deputy Director
4 个月Thx for sharing!
Co-Founder @ Memory Anchor | Changing the Way the World Remembers
4 个月Still need to test this at Arlington ??