Determine ApplicationIt is vital to select a correct Piksi™ receiver and determine your application needs before moving forward. Piksi (v2.3.1) is cheaper, smaller and takes less power but it requires clear open sky view, the operation of Piksi base station, and the device can take from 5 to 15 minutes to initialize to its high precision mode. Piksi Multi costs more and requires more power but it has many performance benefits: faster convergence time, better accuracy on longer baselines and much faster reacquisitions.
Physical IntegrationPhysical integration is one of the most important parts of Piksi deployment. It is necessary to isolate Piksi from shock and vibration and to mount the OEM board far from sources of RF radiation. It is also necessary to mount the GPS antenna with a ground plane and far from any other RF transmitters or noisy electronics.
Electrical IntegrationElectrically, Piksi will require a clean power source. Be careful not to provide power at voltages that exceed the specification. Additionally, the unit will require a means of communicating with a host processor via one of available interfaces: UART, USB, CAN (Piksi Multi only), Ethernet (Piksi Multi only).
Software IntegrationFor most uses of Piksi, one should only use the libSBP serial library and documentation to decode serial data received from the Piksi. It is not necessary to build or compile Piksi Firmware or libswiftnav in order to integrate with the device. The Piksi tools source code repository can provide examples for using Python to communicate with the device, but it is not necessary to install this repository in order to interface with the device.
It is also necessary to determine the host system which will communicate with Piksi. It is typical to use an existing embedded system, flight controller / autopilot, or any on-board processor to communicate with the device. This communication can take the form of reading and interpreting messages to be used in real time or logging data to disk for post-processing.