Positioning Australia for its farming future

Transcription

1 Positioning Australia for its farming future Utilizing the Japanese satellite navigation QZSS system to provide centimetre positioning accuracy across ALL Australia David Lamb 1,2 and Phil Collier 2 1 Precision Agriculture Research Group University of New England, Armidale & 2 CRC for Spatial Information

2 Field surveys Opportunities of precision (today) CTF Asset management ~1-5 m Yield monitoring Zonal management ~50 cm Soil & water management ~2-5 cm Inter-row operations

3 GNSS (GPS) positioning All about ranging!!

4 GNSS (GPS) positioning All about ranging!! ~cm positioning accuracy requires estimating the distances to 2 cm over 20,000 km distance

5 GNSS (GPS) positioning Calculating the distances to satellites Code-based ranging Pseudo-random code (PRC) used to work out distance to satellites (n ~ 1 MHz; l ~ 300 m) Real satellite PRC is noisy!

6 GNSS (GPS) positioning Carrier-based ranging Measuring the length of the carrier wave itself by counting the number of wavelengths ( ambiguities ) on it (n ~ 1500 MHz; l ~ 20 cm)

7 Main sources of error in position solution GNSS receivers assume all signals from satellites travel in direct straight lines, at the expected speed (300,000 km/s), that the receiver clocks (used to sync PRC) are dead-on-time with the satellite clocks, We know EXACTLY where the satellites are

8 Main sources of error in position solution GNSS receivers assume all signals from satellites travel in direct straight lines, at the expected speed (300,000 km/s), that the receiver clocks (used to sync PRC) are dead-on-time with the satellite clocks, We know EXACTLY where the satellites are No, No, No and No!!

9 Approach #1- Differential GPS Radio link to rover Base station (~2-5 cm) eg- Real-time Kinematic (RTK)-GPS Requires base station GPS located at a known location, and a rover to record remote locations. The errors on the known location are used to correct the remote GPS. The accuracy of correction decreases the further the rover is from the base. (1 part/million)

10 Continuously Operating Reference Station (CORS) A network of permanent GNSS tracking stations Unstaffed, permanently configured to continuously collect and record GPS data.. Local correction solutions (differential) are created and transmitted for integration into live correction systems

11 Opportunities of precision (tomorrow) Autonomous systems Signal integrity, safety operational future!!

12 Opportunities of precision (tomorrow) Animal systems

13 Base station RTK Private ~cm positioning today $10k base station + radio transceiver (25 km outermost limit for link plus accuracy) Corporate Cluster/array of base stations (can be average of cluster)- not talking to each other Some are CORS-ready A future where every farm has one is not feasible

14 ~cm positioning today CORS (Network RTK)- relies on mobile modems!! Current correction methods require high-density CORS (~ 70 km max spacing)

15 Approach #2- PPP- Precise Point Positioning Uses CORS + ionospheric modelling to work out precise clocks and orbits before estimating ranges to satellites. Is not a differencing technique- only requires SINGLE receiver

16 The Problem!! We assume industry (both supply and demand) is taking up positioning by itself! Brand-specific quirks have appeared that could limit uptake!! A NPI including nationwide (non-brand specific) delivery platform of correction signals may help.

17 The QZSS PPP system Current LEX message availability with one QZSS satellite Expected LEX message availability with 3 active inclined QZSS satellites in 2018 Quasi-zenith Satellite System (QZSS) 3 satellites- high elevation (good for built environment too) Transmits on extra L-band channel (LEX) good for transmitting PPP correction

18 Correction messages broadcast on QZSS s LEX signal QZSS trial PPP correction messages generated in Japan using global ionospheric model GPS corrections applied on-the-go PPP reference data derived from Australia CORS (but ONLY 7 ref stations)

19 While we were at it

20 Satellite-only QZSS trial fully-robotic cm positioning tractor Jerilderie NSW, Jan cm rear crawlers on 40 cm row spacing RTK PPP-AR

21 PPP Positioning accuracy verified! Cross-wise deviation (m) NRTK Horizontal RMS = 3.4 cm PPP-AR Horizontal RMS = 2.5cm

22 Key research challenges 1. Reduce convergence time- ~ 5-15 mins Develop local PPP solution (around local ionospheric modelling)

23 Key research challenges 2. Receiver design (integrated LEX receiver/decoder) Compatible with contemporary farming machinery AND future devices (eg drones, robots ). Feasibility of Australia vs Japanese build Multi-GNSS capability (eg BDS)

24 Key research challenges 3. Supply chain feasibility PPP-AR signal Receivers/Decoders Education/Extension/Outreach

25