Kamron Ebrahimi (email@example.com)
Blake Hudson (firstname.lastname@example.org)
Grayland Lunn (email@example.com)
Landslides cost the United States an estimated $3.5 billion in infrastructural damage per year. Regularly monitoring the activity of landslides with high spatio-temporal resolution and accuracy can provide valuable early warning information and aid in the interpretation of landslide kinematics. Traditional landslide monitoring techniques often require drilling, which is expensive, time-consuming, potentially dangerous, and limited in serviceability over time. Owing to these constraints, the Slide Sentinel aims to create a low-cost, easily deployable, landslide monitoring system capable of producing centimeter-level positional readings using RTK GNSS.
The system consists of a centralized base station and a network of rover units which may be deployed over the extents of active landslide terrain. Rover units on the network wake at configurable intervals, receive RTK correctional data from the local base station, and produce a positional reading. Readings are sent to the base station and are uploaded via an Iridium satellite link to a Google spreadsheet for real-time updates of rover movement. Rover units come equipped with a 6 millig-precision accelerometer capable of waking the device in the event of sudden movement and can be remotely configured from a client portal to increase wake frequency and verbosity.
Use RTK GNSS to produce centimeter-level positional readings of rovers on the network
Produce positional readings every 15 minutes
Survive in challenging environments for upwards of 3 months with a low-power efficient design
Produce a scalable network in which a single base can service 6 rovers over an RF link
Configure rover verbosity and wake interval remotely via Iridium Satcom link
The first GNSS receiver used in the design was the Skytraq S2525F8-GL-RTK. While extremely low cost (~$70), this receiver required 30 minutes to produce a reliable, accurate positional reading. Over the course of this wake interval a substantial amount of power was consumed by the RF link, severely infringing on the devices power budget. The device also performed poorly in 15m canopied environments, suffcient grounds to halt development against the unit. While this receiver is not a viable solution to for rover units deployed to canopied environments, or rovers which need to wake frequently, this receiver would be ideal for a cheap, open-sky monitoring solution with 3 to 4 hour positional readings.
While the Skytraq S2525F8-GL-RTK did not provide the required performance for the Slide Sentinel, field testing this receiver provided the team insight into what makes RTK possible and the different factors one must consider when selecting a GNSS receiver for high multi-path environments.
Recent investments in satellite infrastructure by the United States (GPS), the EU (Galileo), China (Beidou), Russia (GLONASS), Japan (QZSS) and Indian (IRNSS) have made multi-constellation, multi-frequency GNSS possible. Dual-frequency GNSS is the most recent development to hit the low-cost GNSS receiver commercial market, largely carried by the Swift Piksi Multi and Ublox ZED-F9P.
Satellites are capable of transmitting duplicate data across separate bands. Satellite within the United States’ Global Positioning System (GPS) transmit on the L1 (1575.42 MHz), L2 (1227.6 MHz), and L5 (1176.45 MHz) bands. While GPS satellites may transmit the same data on each of these frequencies not all satellite antennas or GNSS receivers may be capable of receiving and performing corrections using the data carried on these separate bands. GNSS receivers capable of receiving signals carried on two frequencies of different constellations are called dual-frequency GNSS receivers. Dual frequency satellite reception allows for the robust correction of ionospheric attenuation by the receiver. While a single-frequency GNSS receiver might produce positional readings within 5 meters of the receivers actual location, a dual-frequency GNSS receiver can provide readings within 30 cm.
In addition to compatible reception frequencies, GNSS receivers also differ based off the number of navigational satellite constellations they can receive and process data from. A receiver capable of processing signals from a wide range of satellite constellations can effectively “see” more satellites for better positional fixes and lower GDOP. RTK initialization requires a minimum of 5 satellites in view, but the more satellites available the better. As an example of GNSS receiver frequency and constellation specifications, the Skytraq S2525F8-GL-RTK is a single-frequency GNSS receiver capable of only receiving satellite signals from the GPS L1 and GLONASS L1. The Swift Piksi Multi is dual-frequency and can process signals from the GPS L1/L2, GLONASS L1/L2, Beidou B1/B2 and Galileo E1/E5b. High end OEM GNSS receivers such as the Trimble BD990 are often triple-frequency and are capable of process satellite information from the following constellations on the following frequencies:
GPS: L1 C/A, L2E, L2C, L5
BeiDou: B1, B2, B313
GLONASS: L1 C/A, L2 C/A, L3 CDMA14
Galileo: E1, E5A, E5B, E5AltBOC, E614
IRNSS: L5 – QZSS: L1 C/A, L1 SAIF, L2C, L5
SBAS: L1 C/A, L5 – MSS
MOVING TO A MODULAR SOLUTION
As the Slide Sentinel has matured and garnered more interest, the team has identified a need for a modular plug and play design. The Slide Sentinel is being developed to support a wide suite of GNSS receivers and communication links for better tailored monitoring solutions.
The two primary deployment opportunities for the system include a heavily canopied active landslide in a remote location and an open sky deformation on the Oregon coast with available cell service. The requirements and location constraints for these two deployments has provided ample reason to keep the design flexible as there is no need to spend excessive money on a triple-frequency GNSS receiver specifically constructed to perform well in tough environments when a low cost dual-frequency receiver would work equally well.
The rover PCB comes equipped with a high-accuracy RTC, the DS3231, a selectable RS232 interface for communicating with either a Freewave Z9-T or Z9-C radio, the Adafruit MMA8451 accelerometer breakout, a tunable voltage divider for reading input battery voltage, and two Polulu high-efficiency regulators. The system is designed to operate with the output of any 12 volt battery or solar charge controller, and exposes a simple interface for communicating with just about any GNSS receiver. The design utilizes an SPI low side driver, the MAX4280, for easily switching power to the radio and attached GNSS receiver while conserving pins. The PCB allows for the easy insertion of the Adafruit Adalogger, which comes equipped with micro SD logging capabilities for saving all positional readings. You can download an in-depth build guide for the PCB under “Resource List”.
One challenge encountered by the team was the need to pipe both correctional RTK data coming from the base station to the rover GNSS receiver and also route any configuration data (verbosity and wake interval) to the rover microcontroller. to accomplish this the PCB comes equipped with the SN74LVC2G53, a single pole analog switch. This switch coupled with a simple flow control protocol, allows rovers to request correctional data from the base station, receive any queued configuration information from the base, then switch incoming RTK correctional data from the radio to the rover’s GNSS receiver.
Development on the project is currently focused on reducing rover power consumption, improving network scalability and researching commercially available OEM GNSS receivers capable of getting RTK fixes in canopied environments.
The final goals of this project is to design a system in which a single base can service up to 50 rover units. Rover units will upload positional information to a database where the positional readings will be displayed in real time to a client web app rather than a Google spreadsheet. All information pertaining to individual rovers health will be easily and intuitively displayed.
de Marsily, G. (1986). Quantitative hydrogeology: Groundwater hydrology for engineers.San Diego, CA: Academic.
Klotz, S., & Johnson, N. L. (Eds.). (1983). Encyclopedia of statistical sciences, Hoboken, NJ: John Wiley.
Tapley, B. D., & Kim, M.-C. (2001). Applications to geodesy. In L.-L. Fu & A. Cazenave (Eds.), Satellite altimetry and Earth sciences: A handbook of techniques and applications(pp. 371–406). San Diego, CA: Academic.
Khain, A., Pokrovsky, A., Blahak, U., & Rosenfeld, D. (2008). Is the dependence of warm and ice precipitation on the aerosol concentration monotonic? Paper presented at 15th International Conference on Clouds and Precipitation, Cancun, Mexico.