about GPS & INS?

Question

we will be integrating GPS (garming 16-LVS) and INS (ADIS16355) that would be used for real time tracking of a vehicle. do you think this would be enough to provide a good result? Due to limits in budget what suggestions do you have?

Answer 1

Of cource, it depends on the desired accuracy. 1m accuracy might be feasuble in the environment that doesn't have a lot of multipath.

Strictly speaking, what you're building is not INS, but IMU (inertial measurement unit). ADIS16355 doesn't have enough stability (even if you ovenize it) to make an INS, because it's a cheap MEMS.

I'd like to invide your question to the group dedictaed to sensors: http://tech.groups.yahoo.com/group/sensorforum

Answer 2

Are you using the Garmin to re-reference the inertial system every second? Or are you using the inertial sensor to interpolate between fixes from the Garmin?

The thing is, what result are you looking for? We need to know your criteria before anyone can say whether you have a good chance of meeting them.

For car navigation and the like the inertial element is typically very poor, but it is just keying the GPS when it doesn't have enough information to know what's going on, like course changes with little motion and when denied satellite coverage.

It's not like you are trying navigate precisely in 3-D and can only update the INS reference once in a while.

I think you have plenty of data to derive good tracking, unless the vehicle goes in to a parking garage and drives randomly around for a while.

Pay attention to the pdop data from the Garmin, but otherwise treat its positions as definitive then interpolate between then with the deltas from the inertial sensor.