1. HIPS - UCD : Localization and Orientation Capabilities of GPS HIPS Project Meeting IRST, Trento, Italy 28 - 29 April 1998 G.M.P O’Hare, R. O’Rafferty, K. Sewell

2. Contents Garmin GPS II+ description Connecting Apple MessagePad with Garmin GPS II+ Field Trials with Garmin GPS II+ Conclusions and Recommendations

3. Garmin GPS II+ Specification (1) 12 Parallel Channel Receiver – Can track up 12 GPS satellites at once Accuracy –1-5 m with DGPS –Only 100m is guaranteed due to SA Physical Specification –Size : 5.86 x 12.7 x 4.11 cm –Weight : 255 g

4. Garmin GPS II+ Specification (2) Initial Position Acquisition Time –15sec (Warm start) –45sec (Cold Start) –Position is updated every 1 second Interfaces –NMEA 0183 protocol is output over RS-232

5. Garmin GPS II+ Specification (3) Navigation Features –GPS II+ can store the following 500 waypoints (an individual Latitude/Longitude reading) 20 Routes containing 30 waypoints each –These pieces of data can be uploaded to a PC/Mac using Garmins GRMN/GRMN protocol

6. Connecting Apple MessagePad with Garmin GPS II+ (1) Physical Connection –Garmin Serial cable connecting data output port of the GPS II+ to the Serial port of the Apple MessagePad (RS-232 connection) GPS II+ Data Output –The GPS II+ outputs data complying with the NMEA (National Marine Electronics Association) 0183 standard at 4800 baud

7. Connecting Apple MessagePad with Garmin GPS II+ (2) Sample NMEA 0183 Sentence –$GPGLL,53.18.595,N, 006.13.400,W –GPGLL : Indicates that the data was generated by a GPS device (lat/long) –53 degrees, 18’,595” North –6 degrees,13’,400” West –Sentence is terminated with Carriage Return, Line Feed

8. Connecting Apple MessagePad with Garmin GPS II+ (3) Software Interface –Written in NewtonScript using the Newton Communications API, the “Endpoint” –Data bits, Stop bits and bps options are set to Interoperate with the GPS device

9. Connecting Apple MessagePad with Garmin GPS II+ (4) PROCEDURE MakeConnection() Begin EndPoint:Instantiate(Endpoint Options); EndPoint:Bind(Serial Port); // Bind code to Serial Port EndPoint:Listen(); // Wait for Input EndPoint:Accept(); EndPoint:ProcessInput(); // Parse the NMEA data EndPoint:Disconnect(); End

10. The IMP TourGuide Field trials were carried out using an IMP (Intelligent, Mobile, Personalised Systems) demonstrator, which is a handheld tourguide for the UCD campus The IMP TourGuide provides two modes of operation –Static Tour –Dynamic Tour

11. Static Tour –User makes initial exploration of location by interacting with a map displayed on the MessagePad –When a location on the map is clicked, relevant information is presented to the user The IMP TourGuide (2)

12. The IMP TourGuide (3) Dynamic Tour –The dynamic tour provides tour information while the user is physically exploring the UCD campus –This mode is pro-active and does not require user interaction to function

13. The IMP TourGuide (4) Dynamic Tour –Positional information is captured from the GPS device –Location relevant information is retrieved using the position data (lat/long) as the lookup key

14. The IMP TourGuide (5)

15. The IMP TourGuide (6) While the user is taking the tour, it is possible to build up a route and store this information along with the users profile The Garmin GPS II+ currently allows the user to upload waypoints and route to a PC / Mac after usage (effectively offline) Another option is to store the positional data on the palmtop, such that the route is stored when the tour is finished and can be automatically uploaded to the database

16. The IMP TourGuide (7) PROCEDURE StorePos(Lat,Long) BEGIN // Store Current Position in Route route[next].lat := Lat; route[next].long := Long; route.next := route.next + 1; END

17. Field Trials Field trials were carried out using the IMP TourGuide Experiments focused on the following: –Position Lock –Affects of building coverage –Accuracy –Orientation

18. Field Trials - Localization (1) Position Lock –A GPS signal cannot be received indoors –Require a minimum of 3 GPS signals (2-D) –3-D Navigation (lat/long/alt) requires at least 4 GPS signals –Initial usage requires 15-45s the find position Building Coverage –Tall buildings can affect the GPS signal if the user is less than ~5m from the building

19. Field Trials - Localization (2) Building Coverage –may lose one GPS signal –Altitude parameter may be lost (3D -> 2D) Accuracy –Given the situation with Selective Availability (SA), Accuracy can be off by up to 100m –Trials with the GPS II+ set accuracy at ~15m

20. Field Trials - Localization (3) Accuracy - DGPS Add-on –An accuracy of 1-5m can be attained with a Differential GPS (DGPS) beacon –DGPS is removes SA error with a different thirds party signal –DGPS costs twice the price as a GPS receiver –DGPS requires a 1-2m antenna

21. Field Trials - Orientation GPS devices do not have Compass a capability Heading can be calculated given 2 positions Heading, in general is only useful for in-car usage, or user with definite current heading Without a current and constant heading Orientation is inaccurate

22. Conclusions and Recommendations Outdoor Localization is possible with an accuracy of ~15m Signal maybe lost for short periods in areas of heavy building coverage A PCMCIA GPS maybe more appropriate as the GPS II+ is bulky

23. Conclusions and Recommendations With an on-foot tourist in mind, orientation is not reliable from a GPS device. Electronic compasses maybe an option –Precision Navigation Inc