There is some fantastic expertise in the GNSS hardware sector who know the technical specifications of their products to an amazing level. And while we do know the concepts of location products to a great level, TerraLab are not that. What we can offer is a wealth of experience in using these products in different industries and different real world applications. By understanding your organisational needs, we can guide you through the pro’s and con’s of your available options, which includes price.
Definitions
First of all, lets address the elephant in the room. We call it GNSS, or Global Navigation Satellite System. You might call it GPS, or Global Positioning System. We’re talking about the same thing. The reason we prefer the term GNSS over GPS is because GPS is a specific system operated by the USA, whereas, GNSS encompasses a range of systems. They’re described below:
Global systems:
GPS (Global Positioning System) – USA
GLONASS (Global'naya Navigatsionnaya Sputnikovaya Sistema) – Russia
BeiDou – China
Galileo – European Union
Regional systems:
NavIC (NAVigation with Indian Constellation) – India
QZSS (Quasi-Zenith Satellite System) – Japan
So, when we say GNSS and you say GPS, we probably mean the same thing. But we’re going to keep saying GNSS here.
GNSS for GIS
Accuracy
Today GNSS is ubiquitous found mostly in mobile devices, but also commonly used in industry for tracking, wayfinding or reporting locations. This could be anything from tracking the location of a bus on a transport network, to a surveyor finding property boundaries. The accuracy requirements of each are vastly different. On one hand, the bus needs metres of accuracy, whereas, the surveyor needs centimetres, or even millimetres of accuracy. And the hardware that each use is appropriately scaled in price, size, complexity of use, etc.
So it’s fair to say that for many real-world applications, a GIS professional does not necessarily need the same equipment as a surveyor. Or if they do, it wouldn’t be every day. We classify GNSS hardware into the following categories:
Consumer grade GNSS, operates in the 1.5 m to 5 m horizontal accuracy range.
GIS grade GNSS, operates in the 0.5 m to 1.5 m horizontal accuracy range. And
Survey grade GNSS, operates in the <0.5 m horizontal accuracy range (but usually in the <0.05 m range).
So based on that, we consider most GIS needs as exceeding that capable of consumer grade products, but not necessarily always needing survey grade equipment.
Methods of acquiring sub-metre location
Related to accuracy is how your GNSS gets its position. There are a range of options available including:
Differential GPS: This is a grab-all term which refers to a system that adjusts or recalculates its reported position from the satellites, based on communication with a ground-based reference station. It encapsulates some of the methods listed below.
Real-time Kinematic (RTK): This is a form of correction that surveyors often use (which requires specialist equipment). It uses information from the GNSS signal and data from a reference station to provide real time corrections in location. The corrections are usually transferred from the reference station to the rover via the internet or radio transmission. The accuracy of RTK can be millimetres.
Post-Processed Kinematic (PPK): This is similar to RTK, except that the raw data is kept and calculated after mapping. It might be used in situations where corrections are not possible to be obtained in real time, but the accurate location of a feature needs to be recorded. It is not useful if your accurate position is required in real time.
Precise Point Positioning (PPP): This method uses near consumer grade GNSS equipment, a data connection and more processing power to provide up to a few dozen centimetre position. While the GNSS can be at the lower end of the spectrum, the processing involved does mean a capable device needs to accompany the GNSS.
Satellite Based Augmentation System (SBAS): This method is similar in concept to RTK, but vastly different in practice. It also uses a reference station, but instead of the correction being sent directly to the rover, it is sent back up to the satellites, which transmit the corrected signal alongside the original signal. Because multiple signals are involved, your GNSS needs to be capable of receiving the second signal. The level of accuracy varies based on the specifications of your device but can range from a few dozen centimetres to 1.5 metres.
Other corrections: There are other options available, such as other methods of satellite corrections using L-band correction services.
Each of these methods is capable of achieving sub-metre accuracy if set up correctly and used in the right situation.
Ease of use
A significant factor in GNSS equipment is its ease of use. Quite often a mapping project will be for GIS reasons, but will not utilise the services of a GIS professional. For example, a municipality may require an update to their city infrastructure mapping, but it would take the GIS coordinate years to complete on their own, so they enlist the services of the assets team who may not have the same level of technical expertise. Therefore, the equipment needs to be easy to use.
The ability for a user to pick up a device and connect it to their mobile device is an extremely valuable workflow because 1. They probably already have a mobile device, 2. They will have a data connection to that mobile device, 3. They already know how to use that mobile device, and 4. Using a separate GNSS receiver means they have an independent battery which won’t drain the battery of their mobile device.
These days, this solution is usually a separate GNSS with a reliable Bluetooth connection and an easy to use app.
Size
As you go from consumer grade GNSS, to GIS and survey grade GNSS, you are probably going up in size and weight. And while you are getting a better product (for example, better accuracy), you need to consider if that is practical for the user to be carrying it all day.
Sure you can mount it in a backpack, but at that point are you really getting the accuracy you paid for? If you have expensive survey grade GNSS equipment and you’re not using a survey pole and level, you’re not getting the centimetre accuracy you paid for. Hovering your backpack over a location to be mapped is not centimetre accuracy. If a backpack is your workflow, then you certainly want a GIS grade GNSS.
Flexibility
Sometimes your needs might change. One day, ease of use may be the most significant factor, where as on another occasion, accuracy might be important. While it might be possible to have different GNSS receivers for different situations, wouldn’t it be great if you could just have one? Well, those options exist. But you may sacrifice something like size or cost to achieve that flexibility.
Reliability of signal and your environment
This one is important. The options for sub-metre accuracy for a farmer in a field are different to those of an asset inspector in a city, or a forester in the wilderness. Some methods of correction are so fleeting, whereby, small obstructions caused by trees or buildings may block the signal and force the hardware to take the next few minutes to reacquire the signal. This is not a great solution in a setting where you are locating hundreds of positions in a challenging environment, but may be ok if you are only trying to locate a limited number of locations.
Others, such as RTK over NTRIP (RTK corrections that come via the internet) are dependent on mobile/cell service. If you lose signal, you lose the ability to map at centimetre accuracy.
Cost
Cost is a huge factor for any organisation. Often, all factors mentioned above are improved by increasing cost but not all teams have deep pockets. Or maybe you need multiple devices which obviously multiplies the investment cost.
Our advice is to pick the elements that are important to you and purchase accordingly. If your budget is limited, it is unlikely that you will satisfy all of the items listed above. A ballpark for costs is that you’re going to pay anywhere from $1,000 to $20,000 (AUD) for a GNSS reciever.
Related to cost is also subscriptions. There are many service providers out there that offer correction services for their hardware at a price. The accuracy of the service may be great, but a subscription may not be a great solution for a team that doesn’t utilise it often.
Rent versus own
The final aspect we want to highlight is the rent versus own debate. If you do not use the mapping hardware often, then perhaps renting is the way to go. By renting, you often get the latest hardware, support, cost certainty and no commitment to exorbitant costs or subscriptions. On the flip side, if your project is ongoing or you utilise your hardware frequently or unexpectedly, then purchasing might be the way to go.
Our experience with some GNSS receivers
Below are some examples of GNSS hardware that we have used, how we used it and what we found. We’re not advocating for these in any way, nor are we suggesting we did a thorough review. We’re just listing the details in an easy to read format. Therefore, the discussion is lacking a lot of technical detail. The accuracies we list are horizontal (not vertical) and from a rapid appraisal of GPS metadata from our projects, not from any formal statistical analysis.
Hemisphere DGPS using the Australian Maritime Safety Authority (AMSA) differential correction service. This unit was used in environmental monitoring to mark scientific sample locations. It is an older piece of hardware and now uses a discontinued service that was shut down due to the increase in competing, and frankly better, technologies. It consistently and reliably got accuracies of <1.5 metres prior to an SBAS being available which was satisfactory for its time, but is no longer a suitable option.
Bad Elf Surveyor using SBAS. We used this during the SBAS test bed phase (~2018) to test the value SBAS offered to cheaper, single band GNSS receivers. We tested it in a range of situations and noted that it took a few minutes to get an SBAS correction when first turned on, and lost the SBAS connection pretty easily under trees (but was rapid to reacquire it). Nevertheless, it’s price is the second lowest of all the options listed here. We consistently got accuracies <1.3 metres and got as good as 80 centimetres.
Emlid Reach RS2 as an RTK rover. We received RTK corrections via the AusCORS NTRIP broadcaster. This was used in a range of settings including asset mapping, underground services mapping, and environmental monitoring. While we consistently reported a 1.2 centimetre accuracy, we found the unit large and heavy as the battery and antenna are integrated and located at the top of the survey pole. It was fine for short mapping stints, but in one situation where we were mapping all day with a survey pole, we found it to be very taxing. In some situations we mounted it in a backpack which eliminated the value in such a highly accurate device. Though we did find it to be incredibly reliable and easy to use, so long as an internet connection was available to receive the corrections. If you have two and the know-how, you can set these up to work without mobile/cell signal in a base and rover situation.
Custom RTK setup using u-blox module. Similar to the Reach RS2, we received RTK corrections via the AusCORS NTRIP broadcaster and we used this device for cultural heritage mapping and environmental monitoring. This option was the gold standard for price being the cheapest of the lot, as it was made from modular parts and some expertise. It was also extremely small and light weight being smaller than a mobile phone (except for the pole and antenna of course). The issue was its complexity of use. Because it needed know-how to put together, it also needed know-how in operating it effectively meaning it wasn’t a great solution for other team members without rigorous training. Because of this, when technical problems were encountered, it was up to us to solve them. We less-than-reliably got 1.4 centimetre accuracy from the hardware.
EOS Arrow Gold. This device was used for city infrastructure mapping using RTK corrections via the AusCORS NTRIP broadcaster and remote environmental mapping using the Atlas H50 correction service. The units are feature rich, and provide a fantastic, flexible option for users. In the RTK scenario, it consistently reported a 1.2 centimetre accuracy but we had frequent connection issues to our mobile device. This was eventually overcome but it did make us a little weary of dropouts in the future. While the Atlas H50 scenario was great for our remote setting, it was not a sustainable solution for our project because key satellite providing the service was continually occluded by obstructions causing drop outs and delays. We knew this was a limitation at the onset of the project so were able to persevere, but it is a significant factor for this type of technology. Nevertheless, it is one of the few options listed here that would have even been capable of providing that level of accuracy in such a tough setting. It reported accuracies between 60 centimetres and 90 centimetres in that project. Overall though, the fact that you can use RTK, subscribe to Atlas H50, utilise SBAS, etc. all from one device is an unmatched level of flexibility from one device.
EOS Arrow 100 using SBAS. In similar fashion to the Bad Elf Surveyor, we used this device during the SBAS test-bed for cultural heritage mapping in a suburban setting. It took a few minutes to acquire the SBAS signal, but once acquired we got a reliable signal reporting about 45 centimetres accuracy. We would be curious to see how reliable the SBAS signal is in environmental monitoring settings where occlussion by trees is likely as this unit is much cheaper than the EOS Arrow Gold.
Emlid RX as an RTK Rover. This device is being used in environmental monitoring, and infrastructure mapping. It is designed specifically for obtaining its RTK corrections via the internet (in our case, the AusCORS NTRIP Broadcaster) and is therefore only suitable for mapping where there is mobile/cell service. Because of this, it is much smaller and lighter than its big brother, the RS2 making it a much more mobile friendly product. The app and connection is easy to setup and use. We consistently get accuracies of 1.3 centimetres.
Trimble R1. We used this for environmental monitoring, and although it is SBAS capable, we used it before the SBAS test-bed was operational. We had significant connection and integration problems when first setting up, but once we were up and running it reported 80 centimetres accuracy consistently.
Consider reading this article for a far more detailed test on many of these pieces of hardware.
Closing comments
The range of GNSS devices available for GIS applications is huge. Hopefully we have helped you narrow down your options so that you make the right choice. Feel free to reach out to us if you have any questions - We are not a device retailer so won’t push you towards any particular product. But we will comment on options based on our collective experience mapping for GIS.
Or check out our mapping page if you’d like us to do it for you.
