CanSLAM Consideration: GNSS

What is GNSS?

Global Navigation Satellite Systems (GNSS's) are systems that are essential to every-day global functioning as well as mobile mapping systems and (some) SLAM systems. While you might know the United States "Global Positioning System" (GPS,) it is joined by the Russian Federation (GLONASS,) the People's Republic of China (BeiDou), the European Space agency (Galileo,) and others. Most of these systems (and all listed above) are available for commercial use. While all similar, each system has its unique equipment, algorithms and performance characteristics.

Major GNSS Constellation Origins (EOS, 2024)

Acting in Pairs

While positional data for your phone or car is accurate to within a few metres (3+ feet,) higher performance mobile mapping and SLAM device performance sometimes requires confidences of one to three centimetres (0.4 to 1.2 inches.) For those equipped with GNSS receivers, SLAM and MM systems don't just use their own GNSS antenna. Either in real-time or after the data recording session is complete, information from any and/or many other unmoving GNSS stations ('base station(s)) is incorporated into the system to improve the certainty of the moving antenna (the 'rover') position.

In Post-Processing Kinematics (PPK) GNSS, the rover and the base station collect and store GNSS data but do not talk to each other directly. In 'Post-Processing,' the data stored by one is unpacked and compared against the data stored by the other. Using this data and the positional certainty of the base station, corrections are applied for factors like atmospheric interference and signal 'echoes' (multipath) is filtered further. Location determination is further enhanced this way through infrastructure support services like CSA and NASA corrections to satellite positions for fluctuating atmospheric drag and the updating of atmospheric records with atmospheric satellite data.

In Real-Time Kinematic (RTK) GNSS, the base and the rover perform the same function as in PPK but in real-time. For this, the rover and the base will use either a cellular phone network or radio transmitter/receiver to keep in constant contact with each other. While radio provides faster communication and an additional baseline measurement, they rely on line of sight and so base stations must be strategically located and moved/networked to keep in contact during collection. Cellular RTK connections allow for flexibility through 'over the phone' communication with the trade-off being a delay in communication.

Cellular communication allows for two or more GNSS receivers to talk to each other over longer distances without having to double-back for kilometres to get your unmoving ("base") station. Of course, moving farther away from your base reduces accuracy, so it's necessary to switch or move base stations as ranges reach operational tolerances.

GNSS in Mobile Mapping and SLAM

GNSS is important to mobile mapping systems in particular as they rely on their location data to help determine where they are, what movements they'd gone through, and what the statistically likely paths of travel might be during post-processing. Mobile mapping (MM) systems typically rely on this data, meaning bad GNSS determinations could throw off location and elevation data! By how much? That depends on the other systems that support your scanner and its processes.

Mobile mapping (MM) location accuracy is greatly affected by GNSS position quality. The lesser the quality for longer, the worse the positional certainty.

While relatively simple to use, considerations for accurate GNSS data collection and use are high! Measurement quality factors include aspects like:

• Transmission channels available for use,

• Number of satellites available and in line of sight during the time of collection,

• Arrangement of those satellites when they're there (geometry,)

• Antenna type used to collect the GNSS signals,

• Location of 'phase center' (processing center) in relation to antenna,

• Atmospheric conditions at the time of collection,

• Space weather (yes, it's a thing!) at the time of collection,

• How much of the data packets are retained by the receiver,

• Data packet formatting,

• How much echo or interference there is with the satellite signals,

• How much atmospheric drag is being applied to the satellites vs. expected,

• Atomic/terrestrial clock synchronicity considerations (theory of relativity applies),

• How far a GNSS receiver (rover) moves away from its sister sensor (base,)

• Height reporting type (mathematical ellipsoid, gravity-derived geoid model, or another)

• Number of observations taken.

While SLAM scanners may treat GNSS as 'less trustworthy',' poor location reporting can still throw off sections of a data collection session.

Giving mobile mapping systems in particular such a large 'location' within that range of 'certainty' (uncertainty, really) means that the computer processes are allowed to use that whole area as a 'good enough' place for its trajectory calculations to end up. This becomes particularly problematic when you're trying to accurately reconstruct the estimated trajectory ('travel path') of the instrument and, therefore, the scan it collects.

Unfortunately, this cannot be talked about more without discussing the other instruments that make up a typical mobile mapping unit; the inertial measurement unit (IMU,) the distance measurement unit (DMU,) and the SLAM processes that support LiDAR and/or image-based platforms

There's a lot that applies to and can go wrong with GNSS, which is why it's important to have a geomatics professional on hand!

Sky plot chart for assisting in positional quality determination