Jump to content

 
Photo

GPS accuracy for navigation application

- - - - -

  • Please log in to reply
7 replies to this topic

#1
Nouman

Nouman

    Newbie

  • New Member
  • Pip
  • 3 posts
  • Pakistan

I have a related question to this ongoing topic. So thought of dropping it here instead of starting a new topic.

I am planning to develop a prototype for a GPS (in-car) navigation device to aid in planning a large-scale navigable database development that would be suitable for vehicle navigation, tracking and various location based services.

My question is that while creating the base-map and POI layers how much accuracy is considered good enough for navigation applications. Sub-meter ? I really don't have an idea right now. What my plan for the prototype is to develop base-map and POI layers for a small region and test its suitability by putting it to use with a simple navigation application. I would appreciate any input on my planned work-flow as well. I will be developing the base road netwrok map using satellite imagery and GPS field surveys to collect POIs etc.

#2
Nouman

Nouman

    Newbie

  • New Member
  • Pip
  • 3 posts
  • Pakistan

Hi,

I am planning to develop a prototype for a navigation application in order to aid in planning the development of a large-scale navigable database suitable for in-car navigation applications and various location based services. But before that I have to first develop sample navigable map data of a small part of that region so that I can use it with the navigation application prototype.

For developing road network base map, I am going to use satellite imagery and GPS field surveys for adding POI layers. Can anybody provide guidance in what sort of accuracy is required for navigable maps and what techniques are most suitable for it. The research I have done so far brings out 2 techniques: post-processing and RTK. But I haven't found any mention of them being useful in creating navigable road maps.

Thanks,
Nouman.

#3
frax

frax

    Hall of Fame

  • Associate Admin
  • PipPipPipPipPipPipPip
  • 2,308 posts
  • Gender:Male
  • Location:Stockholm, Sweden
  • Interests:music, hiking, friends, nature, photography, traveling. and maps!
  • Sweden

I think it would matter if it is urban or rural (primarily). I can't imagine it being so important with sub-meter accuracy, especially for POI's - but I guess that would also matter what kind of POI it is (school/gas station vs parking meter).

(I don't have any experience in this, other than as a user...)
Hugo Ahlenius
Nordpil - custom maps and GIS
http://nordpil.com/
Twitter

#4
Hans van der Maarel

Hans van der Maarel

    CartoTalk Editor-in-Chief

  • Admin
  • PipPipPipPipPipPipPip
  • 3,881 posts
  • Gender:Male
  • Location:The Netherlands
  • Interests:Cartography, GIS, history, popular science, music.
  • Netherlands

My gut feeling says to do it as accurate as you can reasonably can. But really it does kinda depend on the local geography (and let's face it, if you're navigating to say a gas station and you end up 10 meters away, that's close enough, isn't it?) and the positional accuracy of the GPS units you'll be using for navigation.
Hans van der Maarel - Cartotalk Editor
Red Geographics
Email: hans@redgeographics.com / Twitter: @redgeographics

#5
Greg

Greg

    Master Contributor

  • Validated Member
  • PipPipPipPip
  • 214 posts
  • Gender:Male
  • Location:Ontario
  • Canada

Most consumer level GPS units have range from 5 to 15 meter accuracy. Sub Meter accuracy would be over the top for recreational POI's.. but it really depends on the hardware and end user application.
Greg Moore

g r e g @ c a r t o g r a p h i c d e s i g n . c o m
www.cartographicdesign.com

#6
frax

frax

    Hall of Fame

  • Associate Admin
  • PipPipPipPipPipPipPip
  • 2,308 posts
  • Gender:Male
  • Location:Stockholm, Sweden
  • Interests:music, hiking, friends, nature, photography, traveling. and maps!
  • Sweden

It also depends on the latitude. I was suprised at how much faster I got a fix and how much better accuracy I got in Canada (e.g. Toronto 43N), compared to back home (Stockholm, 59N).
Hugo Ahlenius
Nordpil - custom maps and GIS
http://nordpil.com/
Twitter

#7
Richard

Richard

    Key Contributor

  • Validated Member
  • PipPipPip
  • 52 posts
  • Gender:Male
  • Location:Tasmania
  • Interests:Surveying, Bushwalking<br />As a kid mapping has always interested me.<br />Main mapping is for a local rivercare group - Mount Roland Rivercare Catchment Inc.<br />Use Manifold, AICS with MAPublisher and CanvasX GIS and field book and pencil.
  • Australia

There's really no drama with post processing data as opposed to RTK.
RTK is only really required if you need to know exactly where you are NOW - ie in the field.
I would think 1 meter accuracy for this would be more than ample and as others have said - if you're that close you can see it then thats all you need.
A 'lower end' GPS capable of 1 meter or sub 1m would be a suitable option - I wouldn't go for anything more accurate as you may well end up with lots of mssing data due to intereference from tree canopy, buildings etc due to the way these survey accurate GPS's are set up - ie to not accept results that whilst good enough for your particular job don't fall within the accuracies as specified (set up by the user) on the machine. These can be tinkered with, but unless you know what you are doing, then you may get a lot of so called positions which may be just a heap of garbage, and you may just as well have run around with your Garmin 12.
that said the newer GPS with GLONASS can offer better coverage in some 'tight' areas.
If your'e buying or leasing then I'd talk to the GPS bods and seek their recomendations. Dont let them talk you into more than you need though.
It may well be wise to specify an envisaged accuracy of your data if youre concerned about that.
Hope that helps.

#8
ProMapper

ProMapper

    Master Contributor

  • Validated Member
  • PipPipPipPip
  • 216 posts
  • Gender:Female
  • Location:India
  • Interests:promapper@gmail.com
  • India

I agree with Richard about accuracy of 1m being more than sufficient for vehicle navigation. The sub meter or cm level accuracy is required in setting up Survey stations or may be tunneling, pipeline alignment and similar jobs. Also the sub meter accuracy is achieved by taking readings from two or more sets of GPS data providers like GLONASS, GALILEO and GPS.

The accuracy can also be achieved with differential GPS by having a known bench mark to provide corrections. But all these require high end GPS sets which may cost more than $2000 per piece. Any vehicle navigation system costing more than $500 would be difficult to sell as other models are already available for lesser price.

And another cost addition would imply if you want the POIs at sub meter accuracy. So it would be double geopardy or something similar; you pay high for creation of highly accurate map data and then you pay more for GPS sets to see that level of accuracy.

I would recommend that you spend a little money in buying an already existing GPS vehicle navigation and use it extensively. Probably as a user you will gain better understanding of its pitfalls than any technical manual full of scientific gibberish.

Anu
http://www.mapsandlocations.com




0 user(s) are reading this topic

0 members, 0 guests, 0 anonymous users

-->