Abstract

Every possible location on the Earth’s surface has been mapped by coordinates, which makes it possible to express any location accurately. Satellites broadcast to these locations so that any receiver may realize its location. This is the theory behind the Global Positioning System (GPS). The purpose of this project is to build a robot that is capable of navigating itself by taking full advantage of this technology.

Objective

In order to fulfill the objectives of this project, certain requirements must be outlined. These parameters will shape the way this project is carried out, from design decisions to economic constraints. The most crucial objective will allow an operating vehicle to navigate from one point to another using an onboard GPS navigation system. The vehicle should not get stuck or turn over easily. The destination and any needed initializing inputs should be the only data input into the system. The vehicle should be able to traverse multiple types of terrains and navigate the entire course (point A to point B) that has been input by the operator. Simple obstacles in the vehicle's path should be navigable. The overall weight of the vehicle, with all components applied, should weigh less than 10kg (22 lbs.). The overall cost of the project, which includes testing and other incidental costs, should not cost more than one thousand dollars.

Design

Compass

Preliminary Design

Since the Garmin GPS-LVS16 only provides a calculated compass heading output when the unit is moving, a stand-alone compass is needed in order to determine the direction that the vehicle is facing. The initial design involved a PNI V2Xe 2-axis 360 degree resolution digital compass. This compass required a 3VDC input which would have been easily

pulled from the Zilog microcontroller board. The means by which it output data to communicate with microcontrollers is SPI communication protocol.

Problems Encountered

Due to the lack of turning precision of our vehicle, a digital compass with great precision was not necessary. The PNI V2Xe’s output using SPI communication protocol was also determined to involve too much work with software.

Final Design

We chose to use the Dinsmore 1490 digital compass with 45 degree resolution (8-directional). The 8-directional compass provided enough resolution to indicate to the software which direction the vehicle was traveling to assist in navigation. This compass outputs to four general purpose I/O pins of our microcontroller. These bits determine the directional orientation of the compass. This was much easier to implement than the PNI V2Xe digital compass. Figure 2 shows the schematic for interfacing the Dinsmore 1490 digital compass to the Zilog Z8 Encore! microcontroller. The following Figure 1 shows the digital compass mounted to the top of the vehicle without the white plastic enclosure.

Conclusion:

In final tests, the GPS Homing Robot would unreliably progress towards its preprogrammed destination. Since the vehicle relied so heavily on the 8-directional compass, it did not have that great of accuracy as far as heading direction. The vehicle was programmed to travel forwards for three heading directions to reduce abruptness of driving towards its destination. Say if the desired heading was north, the vehicle would travel forwards for north, northwest, and northeast.

Without hazard avoidance implemented, the vehicle could not avoid any hazards and the user would have to follow the vehicle closely and pick it up if collision with an obstacle was imminent. All of the hardware aspects of the project were completed well before the Senior Design Day deadline. The software bugs which needed to be worked out were the most time consuming of any problems encountered. Testing for the project was a difficult and lengthy process. Since the GPS only provided valid outputs while outdoors, the only way we could test and debug was through trial and error outdoors. Not only did it take upwards of 5 minutes for each test (GPS acquire time and setup), but the vehicle motor and compass batteries would run out of energy quickly. For the project to successfully fulfill all of its specifications, more software work would need to be done and tested.