A local affordable Bathymetry as a Service - the hardware

written by: Hessel Winsemius Dr. Hubert Samboko Stephen Mather Olivier Hoes

Measuring bathymetry in rivers, along coastlines and in reservoirs is notoriously costly. Surprisingly in Africa, measurements of bathymetry are often outsourced to international firms that fly in with echo sounders and RTK devices. What if a local surveying expert firm could do this instead, do it at scale, and do it within the local economy? And what would that entail? The OPEN PROFILE consortium, funded by WMO, is working on several pieces of the puzzle to do this: one of these is a very affordable, locally reproducible hardware design, using an Android smartphone for control and off-the-shelf DIY sensors for echo sounding.

In October 2023, the OPEN PROFILE consortium team consisting of Delft University of Technology, Rainbow Sensing, OpenDroneMap and University of Zambia set out to figure out what would be needed to get towards such a service. Let’s call this “Bathymetry as a Service” (BaaS). In a number of blogs, we introduce our solutions that should lead to local service providers or entrepreneurs being able to develop their own BaaS in their own country.? Several technical components are needed to deliver this service to key end users such as irrigation management authorities, hydropower authorities, water management agencies, National HydroMeteorological Services. From our key stakeholders we learned that the following points are important.

• The methods should use hardware that is reasonably affordable and can be replaced without too much effort.
• The components should be able to withstand fast flow conditions and heat.
• In-field methods should be low complexity. Ideally a person locally can perform the survey without attendance of a client representative to supervise. The use of a smartphone is seen as a very promising direction.
• The processing in software should be straightforward. Contour lines are an important end result especially if software will be extended eventually to cases applicable with reservoirs.
• Ideally, a tie line measurement should not be required. In many cases, the banks are not easily accessible, e.g. due to vegetation, unstable banks or wildlife.

In the rest of this blog we focus on the hardware solution. In blogs which will follow we will also focus on data collection, software innovations, and data processing.

Hardware

TU Delft and UNZA co-developed a hardware setup that fulfills the requirements. It uses a simple off-the-shelf fish finder and combines this with geographical measurements of location using a very precise Real-Time Kinematic (RTK) Global Navigation Satellite Systems (GNSS) observation method, using uBlox ZED-F9P dual frequency chipsets for the GNSS part.?

Since a couple of years, these chipsets have become highly affordable and thanks to suppliers such as Ardusimple, highly modular break-out boards make development of new sensing projects relatively easy. The dual frequency nature allows for the use of a base station / rover combination resulting in typical horizontal accuracies of observations within a couple of centimeters. The fish finder sensor was tested to be very accurate if used in a small beam mode. We have a nice configuration that replaces your smartphone GNSS positioning with the uBlox chipset’s, hence providing very accurate GNSS positioning on your smartphone. This in turn is used by the fishfinder’s android app to provide a very accurate position on the bathymetric estimates.?

The first difficulty that needed addressing is that fish finders, even though their sensor is very accurate, are not meant to be used in fast moving water or while dragging them through the water. First tests also revealed that the fish finder can lose WiFi connection with the smartphone. Imagine you want to cross a reservoir to measure bathymetry, you want to do it with some speed in order to get your data fast! Or imagine you want to measure in a stream with 2 meter per second flow velocities whilst dragging the sensor from one side to the other on a bridge. Any generally usable solution needs to stabilize the sensor.?

To address these challenges, TU Delft (Olivier Hoes) iterated a number of times with a rig design to provide stable readings and stable telemetry while using PVC plumbing under the premise that PVC and PVC glue can be acquired in most contexts.

The main design requirements for the rig were that it must be easy to assemble, consist of locally available parts, should be stable in high velocity flow conditions, keep electronics dry even if the construction gets submerged.?

The resulting design is small, stable, and has a centrally positioning beam where a rope can be connected. This prevents the rig becoming unstable if it is operated from a higher point (e.g. a bridge) with high flow velocities. The smartphone is carried in a cylinder shaped opening with a lid. The GNSS antenna and radio are carried on top of that. The rig was tested while moving slowly through the water and while moving fast.

After these first designs were made and tested with BSc students in Delft, The Netherlands. Hubert Samboko (UNZA) replicated the rig in Zambia and tested it on a diversion channel of the Lunsemfwa River, validating part availability, local production of the design, and stabilization of the sensor in situ.

A small survey was carried out near the Civil Engineering Faculty of TU Delft, demonstrating the feasibility of carrying out a survey. The approach is straightforward: first you set up a GNSS base station, which first surveys itself, and then continuously carries out navigation satellite data collection to assess the error in positioning. Correction messages are continuously sent to the station on board of the boat. The rig on the boat receives the navigation data from the base station and uses it to perform a very precise positioning with Real-Time Kinematics. Combined with the stable fish finder echo sounder, a precise depth combined with a precise horizontal position is obtained.

The end result: seamless bathymetry of the pond

The conclusion is that we have a solid affordable and easy to establish measurement platform to measure bathymetry up to 100m deep. Price points of the required hardware (base GNSS station, rover materials, fish finder, cabling and PVC materials) is about 1,500 USD only. In addition, the parts are easy to replace if they at some point fail. A typical RTK enabled echo sounder costs in the order of 15,000 USD (excluding a GNSS base station!). With a factor 10 difference in pricing, and a more easy approach to spare parts, we consider our rig design and hardware set up to be very promising for further use anywhere in the world.

The OPEN PROFILE consortium will also post on acquisition of dry bed bathymetry, and software processing methods. These will become available soon with a free and open-source license. Please stay posted.

Hardware guide

We have prepared a hardware guide that allows you to set up a rig and configure and measure yourself. It contains explanations of the required components and how to configure them yourself. All configurations can be done on an Android smartphone and with the U-center application! This makes the hardware simple to use as long as you carefully read our instructions.