Archive

Posts Tagged ‘GPS’

Testing positional accuracy underwater

April 15, 2019 2 comments

I recently had the opportunity to test positional accuracy while underwater.  And I did not even have to get wet!  While I was doing some GIS work with the excellent faculty at the University of Hamburg, I walked through the St. Pauli Elbe Tunnel while collecting a track.  As I did so, I reflected on the fascinating cultural and physical geography of this 1911 engineering masterpiece that is still in use:  The tunnel is 426 m (1,398 ft) long; it was a technical sensation when constructed; photos at the entrance show Kaiser Wilhelm II dedicating it.  It connected central Hamburg on the north side of the river with the docks and shipyards on the south side of the river Elbe.  The most amazing part was the four massive elevators, capable of carrying bikes and whole vehicles, and of course, 100 years ago, carriages and horses.  These elevators and tunnel are still functional and being used today!

While pondering these thoughts, I collected a track in the Runkeeper app, and mapped it as a GPX file in ArcGIS Online as a 2D webmap and as a shapefile in a 3D scene.  I wanted to test how spatially accurate a track underwater would be, in the x and y dimensions, but also in the z dimension.  First, let’s consider the x and y:  As I walked through the tunnel 24 m (80 ft) beneath the surface through one of the two 6 m (20 ft) diameter tubes, I expected the my app to lose sight of the GPS, Wi-Fi, and cell phone towers, but I did not know how far off my position would be.   My recent experiments on an above-ground track gave me a ray of hope that perhaps my position would be recorded as somewhere in Germany rather than in the North Sea or the Atlantic Ocean.

I was told by a local source who said that the tunnels are 8 m below the bottom of the river, making the water 16 m deep here (this depth here allowed Hamburg to become of the largest container ports in the world).  Thus, above me was 8 m of sediment (glacial, in this area), and 16 m of water for a total of 24 meters above me.  The elevation at the water surface here is approximately 5 m above sea level.  Thus, my elevation in the tunnel should be 5 – 24 = -19 meters, minus 3 more meters because I was standing on the bottom of the tunnel rather than the top, so -22 meters.

My results as a 2D webmap and as a 3D scene are shown below.  As is evident, the recorded elevations are all above sea level, at around 4 or 5 meters, so they were 22+5=27 meters off of my tunnel elevation.

elbe_tunnel_experiment_screen

A 2D map in ArcGIS Online showing the results of my experiment, with elevations in meters above sea level shown as labels. 

Feel free to open and interact with the data!  For example, to test the X and Y:  Using the measure tool, measure the distance between the tunnel as shown on the OpenStreetMap basemap and the position recorded by my track.  As I left the train station on the north side, my position was fairly accurately recorded, but once I descended the stairs into the tunnel, my position was off to the east by about 140 meters, and then shifted to the west and was off by about 240 meters.  But as I continued walking south, for the last 1/3 of my trek through the tunnel, my XY positional accuracy was only off by 50 meters.   I ascended the stairs and circled the parking lot on the south side, and was only 1 to 2 meters off once more.  I descended into the tunnel and walked north.  This time, my position was about 100 meters off, becoming worse as I kept walking.  My position overcorrected 80 meters to the north as I ascended the stairs, and “settled back” to being a few meters off as I walked to the train station.

To test the Z position:  The elevations were, as I suspected, not displaying their correct number below sea level; that is, 19 meters below sea level. However, you can see that the elevations are actually quite close to the elevation of the surface of the river in this area; at about 4.5 meters.

elbe_tunnel_gps_3D_scene.JPG

A 3D scene in ArcGIS Online showing the results of my experiment, with elevations in meters above sea level shown as labels and symbolized as cylinders.   Feel free to open and interact with this 3D scene!

Overall, with only a smartphone and a fitness app, displaying the data in ArcGIS Online, I was rather pleased with the fact that my positions all around were usually only in the tens or a few dozen meters off of true. This aligns with my recent reports of above-ground experiments and is further evidence of the improvements in spatial accuracy with all location based services.

Interested in further exploration?  See the evidence of my field trip in the photographs below.

tunnel1.jpgThe enormous elevators that carry pedestrians, bicycles, and even vehicles from the street level to the level of the tunnels.  This one is at the north side of the river with a photo of the opening ceremony with Kaiser Wilhelm II dedicating it.

tunnel2Standing at the entrance to the tunnel; photo also shows one of the glazed terra cotta art sculptures.

Now, go conduct your own accuracy experiments!

–Joseph Kerski

 

Track on Track, Revisited: Spatial Accuracy of Field Data

February 4, 2019 3 comments

Back in 2014, I tested the accuracy of smartphone positional accuracy in a small tight area by walking around a track.  During a recent visit to teach GIS workshops at Carnegie Mellon University, I decided to re-test, again on a running track.  My hypothesis was that triangulation off of wi-fi hotspots, cell phone towers, and the improved GPS constellation would have improved the spatial accuracy of my resulting track over those intervening years.

After an hour of walking, and collecting the track on my smartphone with a fitness app (Runkeeper), I uploaded my track as a GPX file and created a web map showing it in ArcGIS Online.  Open this map > use bookmarks > navigate to the Atlanta and Pittsburgh (Carnegie Mellon University) locations (also shown on the graphic below on the left and right, respectively).   Once I mapped my data, my hypothesis was confirmed:  I kept to the same lane on the running track, and the width of the resulting lines averaged about 5 meters, as opposed to 15 meters on the track from four years ago.  True, the 2014 track variability was no doubt in part because I was surrounded by tall buildings on three sides (as you can see in my video that I recorded at the same time) , while the building heights on the Carnegie Mellon campus were much lower.  However, you can measure for yourself on the ArcGIS Online map linked above and see the improvement over those two tracks taken just 4 years apart.

I did another test while at Carnegie Mellon University–during my last lap on the track, I moved to the inside lane.   This was 5 meters inside the next-to-outer lane where I completed my other laps.  I wanted to see whether this shift would be visible on the resulting map.  It is!  The lane is clearly visible on the map and on the right side of the graphic below, which I labeled as “inside lane.”

To explore further, on the map above, go to > Contents, to the left of the map, and turn on the World Imagery Clarity layer.   Then use the Measure tool to determine how close the track is to the satellite imagery (which isn’t perfect either, but see teachable moments link below).  You will find that at times the track was 0.5 meters from the image underneath Lane 1, and at other times 3.5 meters away.

Both tracks featured “zingers” – lines stretching away from the actual walking tracks, resulting from points dropped as I exited the nearby buildings and walked outside, as my location based service first got its bearing.  But again, an improvement was seen:  The initial point was 114 meters off in 2014, but in 2018, only 21.5 meters.  In both cases, as I remained outside, the points became more accurate.  When you collect data, the more time you spend on the point you are collecting, typically the more spatially accurate that point is.

tracks_comparison

Comparison of tracks taken with the same application (RunKeeper) on a smartphone four years apart illustrate the improvements in positional accuracy over that time. 

To dig deeper into issues of GPS track accuracy and precision, see my related essay on errors and teachable moments in collecting data, and on comparing the accuracy of GPS receivers and smartphones and mapping field collected data in ArcGIS Online here.

Location based services on the smartphone still do not yet deliver the spatial accuracy for laying fiber optic cable or determining differences in closely-spaced headstones in cemeteries (unless a device such as Bad Elf or a survey-grade GPS is used).  Articles are appearing that predict spatial accuracy improvements in smartphones.  Even today, though, I was quite pleased with my track’s spatial accuracy, particularly in 2018.  I was even more pleased considering that I had the phone in my pocket most of the time I was walking.  I did this in part because it was cold, and cold temperatures tend to rapidly deplete my cell phone’s battery (which is unfortunate, and the subject of other posts, many of which sport numerous adds, so they are not listed here).   Happy field data collection and mapping!

–Joseph Kerski

Lasers: The future of data capture and transmission?

December 12, 2016 Leave a comment

Over the last four years we have discussed some of the many challenges posed by the volume of data now available online – issues of quality, determining provenance, privacy, identifying the most appropriate source for particular requirements and so on. Being overwhelmed by the choice of data available or not always knowing what resources are available or where to start looking have been common responses from geospatial students and practitioners alike.

A recent report from the BBC on laser technology highlighted some current and future applications that have or will transform geospatial data capture, including the use of LiDAR and ultra precise atom interferometers that could be used to develop alternate navigation systems that do not rely on GPS. The article also discusses the inherent limitations of our current electronics-based computing infrastructure and the potential of silicon photonics, firing lasers down optical fibres, to help meet the demand for instant or near-instant access to data in the Internet-of-Everything world. If many feel overwhelmed now by the volumes of data available, what will technologies like silicon photonics mean for data practitioners in the future? Just because data may be available at unprecedented speeds and accessed more easily, that alone doesn’t guarantee the quality of the data will be any better or negate current concerns with respect to issues such as locational privacy. A critical understanding of these issues will be even more important if we are to make the most of these advances in digital data capture and transmission.

The National Geodetic Survey Data Explorer and Citizen Science

June 29, 2015 1 comment

The National Geodetic Survey (NGS) Data Explorer is a web mapping application, launched by the Survey in 2012, allowing users to view geodetic control data across the USA and its territories.  To use, zoom in on the map on a location of interest, and select “plot marks”.  You will see all of the control marks in that vicinity, including CORS, GPS sites, horizonal control markers, and vertical control markers.  Furthermore, the NGS datasheet documentation for each control mark is accessible from the same mapping interface, including the latitude, longitude, elevation, position source, complete description of the physical marker, the history of the marker, the condition of the marker, and other information.

The mapping site was launched in 2012 and has seen improvements since then.  I found it easy to use, and very useful. The only thing I could not find that would be extremely helpful is the ability to export from the map and data set to a variety of formats–a geodatabase would be nice, or at the very least, a spreadsheet.  I also could not find how I could “select” points that I was interested in, aside from clicking on each one on the map.

Our book discusses the impact of citizen science efforts on geospatial data.  On this note, the NGS also runs a “GPS on Bench Marks” effort, a citizen science program for finding and reporting on the conditions of NGS benchmarks.  By providing GPS on bench marks today, people can help NGS improve the next hybrid geoid model, increasing access to the North American Vertical Datum of 1988, and enabling conversions to the new vertical datum in 2022.  Participating could also help the local surveying community know about nearby marks by improving scaled horizontal positions and updating the mark condition or description by submitting a mark recovery.  A web map in ArcGIS Online is here.

If you are interested in other activities and services from the National Geodetic Survey, see the recent excellent summary in The American Surveyor.  This includes guidelines for using post-processing GPS technology to establish accurate ellipsoid heights and orthometric heights, the new North American Vertical Datum that will be released in 2022, and updates on the GEOCON datum transformation tools.

National Geodetic Survey Data Explorer

National Geodetic Survey Data Explorer.

Track on Track: Reflections on GPS Accuracy on a Running Track

November 23, 2014 2 comments

Recently, while at the Applied Geography Conference in Atlanta, I decided to test the spatial accuracy of my smartphone’s GPS in a challenging environment–a rooftop running track.  Although located on a roof, the track was surrounded by buildings far taller, and in downtown Atlanta, a location with many other buildings impeding signals from GPS, wi-fi hotspots, and cell phone towers. A further challenge to the GPS positional accuracy was that each lap on the track was only 0.10 miles (0.16 km), and therefore, I would not travel very far across the Earth’s surface.

After an hour of walking, and collecting the track on my smartphone with a fitness app (Runkeeper), I uploaded my track as a GPX file and created a web map of it in ArcGIS Online.  As I expected, the track’s position was compromised by the tall buildings–I only had a view of about half the sky during my time on the roof.  As you can measure for yourself on the map linked above, the track lines formed a band about 15 meters wide, but interestingly, were more spatially precise along the eastern side of the track, where the signal was better, as you can see in my video that I recorded at the same time.

Also, as I have encountered numerous times in the past, a line about 100 meters long stretches to the north.  Rest assured that I did not leap off  the building, but rather, the first point that the GPS app laid down as I opened the doors to walk outside was about a block away.  Then, as I remained outside, the points became more accurate.  When you collect data, the more time you spend on the point you are collecting, typically the more accurate that point is spatially.

Track on Track:  Reflections of GPS Accuracy on a Running Track

Track on Track: Reflections of GPS Accuracy on a Running Track.

Another interesting aspect of this study is that if the basemap is changed to satellite imagery, it appears that the track overlaps the tall building to the west.  Try it, using the map link above.  However, a closer investigation reveals that this is a result of the orthocorrection that was performed on the imagery; the buildings do not appear from “straight overhead”, but rather, they “fall away” to the east.  Turn this into another teachable moment:  Images, like maps, are not perfect, but they are very useful.  We can learn to manage error and imperfection through critical thinking and through the use of geotechnologies.  This is a central topic of our book and of this blog.

To dig deeper into issues of GPS track accuracy, see my related post on errors and teachable moments in collecting data, and on comparing the accuracy of GPS receivers and smartphones and mapping field collected data in ArcGIS Online here and here.

Despite these challenges, overall, I was quite pleased with my track’s spatial accuracy, even more so considering that I had the phone in my pocket most of the time I was walking.

The Internet is dead, long live the Internet

December 30, 2013 1 comment

Almost a year ago we posted a review on the Internet of things, an emerging global network of internet-connected devices and sensors, so with the end of 2013 fast approaching  it seems like a good time to see how things have developed over the last 12 months and what 2014 and beyond has in store for us. In his article How the internet of things will replace the web Christopher Mims predicts that the internet will change beyond all current recognition, with the role of the web reduced to displaying content. Although the dominant ‘species’ of the internet of things is currently the smartphone, with the latest versions kitted out with sensors and apps for tracking and monitoring many aspects of our lives, wearable technology – smart watches, wristbands, glasses, even temporary tattoos – will become increasingly prevalent as personal sensors and the medium for controlling the connected devices around us.

Accompanying these developments in the available devices are significant improvements in the levels of accuracy in location tracking with versions of GPS technology, such as Apple’s iBeacon technology,  that work indoors. With this increasing accuracy comes the emergence of ‘invisible’ or ‘spatial’ buttons, which according to Amber Case (Esri) are simply locations in space in which some response is triggered when a person or a device enters that space. For example, walking into or out of a room automatically turns the lights on/off, or turning on the security system when you leave home. Needless to say, the potential for using this type of technology as a marketing tool hasn’t been missed. British Airways has already launched a new campaign called ‘Look Up‘  with an interactive billboard in London informing passers-by what aircraft is passing overhead and current deals on that particular route.

Along with the changing role of the web, Mims also discusses the emergence of what some refer to as anticipatory computing, as the internet develops from simply responding to requests to anticipating those requests based on past location, actions and preferences. As with most technical innovations, there will be both benefits and costs; the benefits should mean we have much more control over the resources we use, the cost will be having to make a lot of our personal information available to make this happen.

Be Critical of Data–Even when it is your own!

February 3, 2013 3 comments

A theme running throughout our book The GIS Guide to Public Domain Data is to be critical of the data that you are using–even data that you are creating.  Thanks to mobile technologies, anyone can create spatial data, even from a smartphone, and upload it into the GIS cloud for anyone to use.  This has led to incredibly useful collaborations such as Open Street Map, but this ease of data creation means that caution must be employed more than ever before.

For example, analyze a map that I created ) using Motion X GPS on an iPhone and mapped using ArcGIS Online.  It is shown below, or you can interact with the original map if you prefer.  To do so, access www.arcgis.com/home (ArcGIS Online) and search for the map entitled “Kendrick Reservoir Motion X GPS Track” or go directly to http://bit.ly/Rx2qVp.  Open the map.  This map shows a track that I collected around Kendrick Reservoir in Colorado USA.  This map was symbolized on the time of GPS collection, from yellow to gradually darker blue dots as time passed.

GPS track around Kendrick Reservoir

GPS track around Kendrick Reservoir

Note the components of the track to the northwest of the reservoir. These pieces were generated when the smartphone was just turned on and the track first began, indicated by their yellow color.  They are erroneous segments and track points.  Notice how the track cuts across the terrain and does not follow city streets or sidewalks.  Change the base map to a satellite image.  Cutting across lots would not have been possible on foot given the fences and houses obstructing the path. When I first turned on the smartphone, not many GPS satellites were in view of the phone.  As I kept walking and remained outside, the phone recorded a greater number of GPS satellites, and as the number of satellites increased, the triangulation was enhanced, and the positional accuracy improved until the track points mapped closely represented my true position on the Earth’s surface.

Use the distance tool in ArcGIS Online to answer the following question: How far were the farthest erroneous pieces from the lake? Although it depends on where you measure from, some of the farthest erroneous pieces were 600 meters from the lake.  Click on each dot to access the date and time each track point was collected.  How long did the erroneous components last?  Again, it depends on which points you select, but the erroneous components lasted about 10 minutes.   At what time did the erroneous track begin correctly following my walk around the lake? This occurred at 11:12 a.m. on the day of the walk.

This simple example points to the serious concern about the consequences of using data without being critical of its source, spatial accuracy, precision, lineage, date, collection scale, methods of collection, and other considerations.  Be critical of the data, even when it is your own!