[OSM-talk] more accurate open-source gps hardware

[simon at mungewell.org]

> On Thu, 19 Nov 2009, Michael Kugelmann wrote:
>> Alex S. wrote:
>> > http://blog.makezine.com/archive/2009/11/diy_real_time_kinematic_gps.html
>>
>> I would be interested not in a GPS that is very accurate but which is
>> coupled with a gyro to be able to also work in tunnels like an
>> underground line.
>> Did anybody read something like that? I mean: having a small device for
>> affordable price...
>>

The RTK paper does mention use of a 'optional' IMU, so presumably the code
base allows for this. There are really low cost accelerometers/gyros which
could be used.

Again, I believe that the antenna and RF front-end with be the major issue
for a low cost solution.

> My son did some formal work with this, with the accelerometers on and the
> gps
> off, and found that in about 250m he was more than 60km out.


This is the problem with cheap IMU's, their accuracy/offsets pretty
quickly drives the position off if they are not 're-synced' with GPS.

Interestingly a work colleague was telling me about IMUs in running shoes,
which can get 'zeroed' with every stride.... so if you are targeting the
jogging market there is some hope. ;-)

Simon