Submitted to: IEEE Aerospace and Electronic Systems
Publication Type: Peer reviewed journal
Publication Acceptance Date: 3/14/2008
Publication Date: 3/30/2008
Citation: Thomson, S.J., Smith, L. 2008. Crop Dusting Using GPS. IEEE Aerospace and Electronic Systems 23(3):14-17 Interpretive Summary: Global positioning systems (GPS) are essential components of site-specific aerial and ground application systems. In site-specific management (SSM), field inputs such as insecticide or fertilizer are applied in variable amounts only to areas that require them, thereby saving on chemical costs and reducing environmental loading. A remote sensing platform has recently been developed for use on agricultural aircraft. Stand-alone GPS receivers are used to provide location information for images and to trigger cameras automatically when a desired field location is reached by the airplane. Up until now, there has not been a suitable method to test dynamic accuracy of GPS receivers in airplanes that reach typical speeds of 65 m/s. For our study, a novel ground-based triggering device was used to test the accuracy of two GPS receivers used on the airplane. Because these receivers have slow (1-second) position updating, the receivers were evaluated indirectly against a precisely located ground position by comparing their readings with readings from a GPS receiver with faster updating capabilities. This GPS receiver is part of the airplane’s Satloc guidance system and was evaluated first against ground reference location to 0.01-s precision. Tests showed that location readings from the Satloc’s receiver lagged the ground reference point by an average of -4.53 m (S.D. = 0.68) in the east-west directions and led the reference point in the north-south directions by an average of +7.88 m. After this GPS receiver was evaluated, readings from the stand-alone GPS receivers were compared with readings from the Satloc’s GPS receiver. Positioning data from one stand-alone GPS receiver showed a distinct ground speed dependency, making it unsuitable for use in aircraft. Output from the other receiver (designed for aircraft use) was insensitive to ground speed changes but exhibited a look-ahead equivalent to 2.16-s (or 140-m) at ground speeds evaluated. This look-ahead was probably due to the particular GPS filtering scheme used. These potentially large position differences should be considered when using stand-alone GPS receivers for remote sensing in fast-moving vehicles such as airplanes.
Technical Abstract: Global Positioning System (GPS) receivers and GPS-based swath guidance systems are used on agricultural aircraft for remote sensing, airplane guidance, and to support variable-rate aerial application of crop inputs such as insecticides, cotton growth regulators, and defoliants. Agricultural aircraft travel at much greater speeds than ground equipment (65 m/s, typically), so longitudinal (along-track) error in GPS-derived position is likely to be pronounced. Due to rapid ground speeds, determination of exact GPS-derived aircraft location requires special methodologies. To meet the need for determining ground position accurately, a ground-based spotting system using two strategically placed mirrors was designed to reflect sunlight to the belly of the aircraft when the plane passed over a ground reference point. The light beam was detected by a photocell on the airplane, causing a solid state relay to close through appropriate circuitry. When the relay closed, a record was placed in the data file for the airplane’s Satloc GPS receiver, accurate to 0.01 s allowing exact position to be logged. Two stand-alone GPS receivers were evaluated along with the Satloc system, and their readings were compared with readings from the Satloc. For one day of testing used as an example, the Satloc lagged the ground reference point by an average of -4.53 m (S.D. = 0.68) in the east-west directions and led the reference point in the north-south directions by an average of +7.88 m (S.D. = 0.52). Positioning data from one stand-alone GPS receiver showed a distinct ground speed dependency, making it unsuitable for use in aircraft. Output from the other receiver (designed for aircraft use) was insensitive to ground speed changes but exhibited a look-ahead equivalent to 2-s at the ground speeds evaluated.