Networking 3rd Party GPS/GNSS into Lowrance

Ray Valley

Aquatic Biologist and BioBase Product Expert

I frequently get inquiries from current and prospective BioBase users about the accuracy of consumer-grade Lowrance GPS and whether survey-grade 3rd party receivers capable of differential correction (DGPS) or receiving positions from multiple satellite constellations (Global Navigation Satellite System – GNSS) could be used with Lowrance and processed with BioBase.

The first question about accuracy prompted a test in March of 2013 with a Lowrance HDS tested side-by-side with a Trimble GeoXH.  I was pleased to find less than 1m deviation on average from post-processed Trimble DGPS positions.  One meter accuracy and precision is typically sufficient for most boat-based mapping applications. Still, prerequisites for some projects require DGPS, and there are a number of BioBase users who have and still would prefer to have DGPS generated positions to use when logging trips. Thus, I was interested in exploring the capabilities of networking positions from a third-party receiver into a Lowrance HDS.

Test Device: Emlid’s Reach RS

Emlid’s Reach RS, launched in April 2017, made waves in industry because it promised cm-level accurate Real Time Kinematic (RTK) GNSS technology at a very low cost (rover + base system < $1,600 at the time of this writing).  It’s relative ease to operate and wireless configuration/operation through an Android or iOS app offers an attractive solution for environmental and agricultural research.  Further, Reach’s light weight has made it attractive for mounting in drone aircraft and boats.

Reachwebpic
Emlid’s Reach RS+ GNSS Antenna. Users can use 2 antenna (Rover + Base) to enhance accuracy

However, manufacturers of GNSS hardware typically follow construction-industry interface/networking protocols and staying up to date on the latest standards developed for the Maritime industry through the National Marine Electronics Association (NMEA) is likely less of a priority for manufacturers.  As such, most GNSS devices developed for land-based construction or surveying adhere to the NMEA 0183 interface standard. In the Maritime industry (which includes Lowrance and Simrad), this was updated to NMEA 2000 in the early 2000’s with many enhancements in networking capability and two way communication between devices.

Viewing GNSS positions of the Reach Rover on the Lowrance NMEA 2000 Network

Because Reach is not compatible with NMEA 2000, one of the primary efforts in order to make Reach Lowrance/Simrad compatible is to convert the position sentences from NMEA 0183 to NMEA 2000.  Once the position sentences are converted, NMEA 0183 receivers can be connected with Lowrance through NMEA 2000 and the third party receiver can be viewed on the network (e.g., “Device List” below)

20181012_115832
If you successfully convert NMEA 0183 to NMEA 2000 networked to Lowrance HDS, you should see the device listed on the network (see below for steps to network Reach). Emlid’s Forum highlights steps for a Simrad which are almost identical to what we highlight in this blog

 

Configuring the Reach Rover and Base

These are the steps I followed to network Reach into Lowrance HDS:

First, through the Reachview wireless app, I initialized and configured the Reach Rover and Base to send and receive positions.  I followed the wealth of online tutorials and videos available published by Emlid.

Next, in the Reachview app under Position Output, I set serial output baud rate to 4800

SerialOutputImage
Setting positions to feed through Output 1 alone work in some applications, however in my tests with Lowrance HDS, positions would not reliably feed through Output 1.  When I duplicated these settings in Output 2, I had fewer interruptions with position feeds

Finally, I networked the Reach Rover to Lowrance

Accessories Needed:

2Reach-RS-cable-2m-wo-2nd-connector-1
1. Reach NMEA 0183 network cable without second connector. Connect to #2 below (Reach is the “Talker”)
20181024_150829
2. Connect the Reach cable to the AT10 (SimNet) Converter (Part No. 24005936; “Listener”).  Talker wire on the Reach cable was connected to the Listener wire on the AT10.  Ground on the AT10 was connected with a ground on the Reach cable (any ground on Reach works).
SimnetToNMEA2000
3. Convert SimNet to NMEA2000 (Part No. 24005729)
NMEA_2000_setup_simplesetup_zps67e58c5d
4. Connect AT10 to established NMEA 2000 network on vessel. NMEA 2000 starter kits are available at most marine retailers.  This is a helpful video demonstrating how to hook up a NMEA 2000 network.  Image above shows a networked Point-1 GPS.  In the example I describe, the AT10 connected to the Reach Antenna replaces the Point-1
20181004_154634
5. If you have successfully connected and configured everything, you should now see the Reach device (in this example, renamed “REACHROVER”) on the NMEA 2000 network on your Lowrance HDS and Reach positions should feed to the chart display and also to recorded .sl* files

Is Reach RS more accurate and precise than Lowrance GPS?

Once I successfully networked the Reach antenna, then I could get to the important question: Is Reach RS DGNSS more accurate and precise than Lowrance GPS? Of course, the GNSS and RTK capabilities of a Reach Rover and Base configuration should result in more accurate positioning than standard positions from the onboard Lowrance GPS in many circumstances.  However, the minimum precision of positions stored in Lowrance/Simrad .sl* files are 5 decimal places which equates to 1.11 m at the equator and 0.79 m at the 45th parallel. So although Reach may potentially hit the analogous bullseye more often than Lowrance GPS, when it’s off, it can never be closer than ~1 m to the bullseye.

Still, there is often a long way from the theoretical to the actual, and the natural world is often inconveniently variable.  There are several factors that affect actual GPS/GNSS accuracy that bear mentioning.

Prerequisites for high accuracy by Reach RS via RTK requires

  • Clear view by the base station of the sky in all directions visible to multiple satellite constellations.
  • Clear line of sight of the base station by the rover in all survey locations
  • Other geographical or environmental factors that affect satellite availability and reception (e.g., flat terrain, minimal tree cover)

Where RTK is not feasible due to waterbody size and terrain, users can post-process (PPK) their Rover positions with a nearby reference station.  Feeding post-processed positions is currently not an option with BioBase and thus PPK was not a solution explored in this trial.

 

A simple non-scientific trial: Repeated tracks with Reach RS and Lowrance Point-1 GPS

On the afternoons of October 22nd and 23rd 2018, on the St. Croix River in Bayport MN USA, I ran a trial with a Reach RS networked to a Lowrance HDS-12 Carbon. Specifically, I was interested in simulating a common BioBase application: mapping bathymetry in a real-world scenario.  I recorded 3 repeated survey tracks using Reach, and another 3 using Lowrance’s Point-1 GPS (Lowrance’s popular off-the-shelf external GPS receiver). The question I was interested in addressing was whether the Point-1 GPS deviated further on average with greater variability from a target path of travel than Reach RS with RTK differential correction from a base station placed approximately 200 m in plain sight on shore.

The primary challenge of this trial at the outset was that it was on open water and thus impossible to travel over the exact path with each replicate. Although the primary objective was to simulate real-world conditions, I may some winter day repeat this experiment with a tech-loaded sled on snow-covered ice where I have absolute certainty the same path through the snow is traveled repeatedly.

The next series of images along with annotation describe the process of carrying this trial out.

Screenshot_2018-10-23_12.45.50
A route was created on the Lowrance HDS Carbon Chart. The route represented the target path traveled by each replicate of both antennas
20181022_132049
Navigation for all replicates was through the autopilot feature on the Motorguide Xi5 trolling motor.  The autopilot was steered by a GPS within the head of the trolling motor (Pinpoint™ GPS).  At the outset, we expected that inaccuracies and imprecisions due to water movement and positioning from the Pinpoint GPS would affect confidence in any accuracy comparisons between the Reach and Point-1 antennas.  Still, we were interested in identifying any large deviations between the tested receivers
20181022_110344
The Reach RS Rover was placed over the top of the disconnected Point-1 antenna for Reach trials. The Reach Rover was removed for the Point-1 trials.
20181022_123432
Position of Reach RS Base station on the shore of of the St. Croix River at Lakeside Park in Bayport MN USA (Lat. 45.01682788 Long. -92.77304937, 182.368 m height).  The base position was established with a float point solution. Like many US waterbodies, the shoreline of the St. Croix is wooded. This affected access to GNSS satellite constellations and thus may have affected positioning and accuracy of the base.
SatVisBase1022
Visibility of GPS (G), GLONASS (R), Galileo (E), and SBAS (numbers) satellite constellations to the Reach Base.  Despite the base placed at the highest point on shore, trees likely obstructed the view of all GLONASS satellites.
SatVisRover
In contrast, GPS, GLONASS, Galileo, and SBAS constellations were fully visible and leveraged by the Reach Rover.
GIS Layout Base View
Aerial view of the area where trials occurred (St. Croix R. Bayport MN USA). At the boat launch (NOBASE) 562 m away along a tree-lined shore, the Reach Rover did not acquire base positions.  Once a clear line of sight was established between the Reach Base Station and the Rover, Rover positions were corrected by the base positions.
Screenshot_20181022-114412_ReachView
Rover positions were in a “Fix” solution status during most of the Reach trials.

 

ReapeatedTracksReach
Three replicates were recorded with the Reach RS antenna. Reach positions were were fed to the Lowrance HDS and logged in a .sl2 file. The .sl2 file was uploaded to BioBase (biobasemaps.com) and processed by EcoSound algorithms. The GPS coordinate points were exported and displayed in QGIS 3.2.3.  Using the Distance to Nearest Hub feature in QGIS, the shortest distance from each point to the target track was calculated. The average deviation of Reach RS positions from the target line was 1.32 m +/- 0.012 (SE) during this trial.
Screenshot_2018-10-23_13.36.48
GPS Status from a snapshot during replicates with the Lowrance Point-1 GPS.  Note 10 of the 11 GPS satellites in this instance had Wide Area Augmentation System (WAAS) correction which enhances GPS accuracy.  Horizontal Dilution of Precision (HDOP) was less than 1 and considered “ideal.”  Estimated Position Error (EPE) represents a 50% circular error estimate.  EPE is based on stationary positioning. Point-1 uses heading information to correct and smooth positioning along a traveled (non-stationary) path so actual accuracy may be better than this figure.
ReapeatedTracksPoint1
Three replicates were recorded with the Lowrance Point-1 antenna. Point-1 positions were were fed to the Lowrance HDS and logged in a .sl2 file. The .sl2 file was uploaded to BioBase (biobasemaps.com) and processed by EcoSound algorithms. The GPS coordinate points were exported and displayed in QGIS.  Using the Distance to nearest Hub feature in QGIS, the shortest distance from each point to the target track was calculated. The average deviation of Point-1 positions from the target line was 1.49 m +/- 0.013 (SE) during this trial.  Note the higher density of positions with the Point-1 (10 Hz) vs the Reach (~1 Hz).  The process of converting position outputs from NMEA 0183 to 2000 likely slowed position feeds from Reach.

Some Conclusions

First, we were encouraged that we could reliably connect a NMEA 0183 GNSS device to a Lowrance NMEA 2000 network and feed RTK DGNSS positions into BioBase-compatible .sl* files. This is good news for Engineer firms who have GNSS devices already and would like to use with BioBase or for those surveys where DGNSS is required by the end user of the survey data.  Emlid’s Reach RS+ is a low cost, yet high quality solution for DGNSS needs. Still, converting NMEA 0183 to 2000 is a bit more cumbersome than it needs to be. We look forward to the day when more GNSS devices have NMEA 2000 neworking options.

Second, accuracy (as measured as the average distance to the target path) with Emlid’s Reach RS was statistically better than the Lowrance Point-1 even with a less than ideal study design of using a consumer GPS for navigation of repeated tracks (1.32 m vs 1.49 m respectively).  Still differences between receivers were not large and the Point-1 delivers sufficient accuracy for most types of bathymetric surveys.  The dispersion of positions (standard deviation) from Reach RS was also significantly smaller (0.8) than Point 1 (1.0 m) during these trials.  In other words, although precision is limited to ~1 m at the point level due to .sl* file limitations, precision of all points taken together was significantly better with Reach than with Point-1.

Still, BioBase’s primary strength is leveraging low-cost off-the-shelf technology and automation to produce high quality data and map outputs very quickly with very little user input.  Habitats in waterbodies (e.g., aquatic plants, bottom composition, depth) are quite dynamic and 1 m horizontal error is acceptable level of error for most system-level (e.g., whole lake or pond) surveys. In that case, users will find that the Point-1 GPS is sufficient.

Author: biobasemaps

BioBase is a cloud platform for the automated mapping of aquatic habitats (lakes, rivers, ponds, coasts). Standard algorithms process sonar datafiles (EcoSound) and high resolution satellite imagery (EcoSat). Depth and vegetation maps and data reports are rapidly created and stored in a private cloud account for analysis, and sharing. This blog highlights a range of internal and external research, frequently asked questions, feature descriptions and highlights, tips and tricks, and photo galleries.

Leave a Reply