Autonomous Weather Boat for Yacht Clubs - EV SPRING

by Caelan Taylor

EV Spring

I have always wanted to build an autonomous boat but this had not been supported due to its cost and the time involved. However this got support from my Dad when I came up with the weather boat idea at the beginning of the year as the project I would tackle for Level 3 NCEA Digital Technology. The project completed for level 2 NCEA was a yacht racing start system and I wanted to do a project in 2025 that was yachting focused of which the autonomous weather boat ticks the box. My school was concerned that the project was too ambitious and complex since I planned to have 3 computers having to work together. In fact 3 is now 6 computers working together to handle the application, mainly due to the last minute decision to abandon the VHF communication and to change to using the cellular network. In theory I can now be in London and control SPRING back in NZ. The project will also be used for the NZ Scholarship examination.

Labour Weekend deployment

We have struggled to get deployment testing at a Yacht Club this month due to the recent cancellation of racing. Fortunately I was able to do this deployment at WBBC last Saturday. It is the highest wind SPRING has sailed in to date but it did its task admirably getting itself out on the coarse, holding position around 50m north of the top buoy, streaming back wind information to the start box then returning to the club after cancellation of the sailing after the first race. The boat saw gusts of over 40 knots at times and continued to do its job. I now have to finish the project report by the end of this coming Wednesday for NZ Scholarship. They require the report to be 56 pages in defined font and text size.

The first task for the project was to find a suitable container for all the equipment. Initially the web was searched for suitable model boat plans, of which nothing suitable could be found. Dinghies were then considered as they tend to have a wide beam for their length. The Pram dinghy appeared to have an advantage having a bigger volume to length ratio plus good reviews for stability and towing performance. The decision was to use a half size build of a Bruce Roberts 8 foot Pram Dinghy mainly because the plans provided full size patterns. It was thought that this build would provide sufficient volume for the electronic and power systems needed to be installed. The boat decision has proven to be the correct choice as it has performed much better than expected in various sea states. The changes made to the boat is that it is now powered by 4 electric propellers. These are also used for steering and to hold the boat in a very tight position,

as was shown last weekend in the strong north westerly. The only other modification was to fit a seal and a top to prevent the electronics being swamped.

The dinghy build had its problems. The gunnels and chines are oak and although they were in a steam box for 24 hours and attached using Epiglue they each came off the build frame at least 3 times and sprung across the workshop. Screws had to be used on the transom and bow to help the glue. Hence the name for the boat of SPRING.

How it works

A lot of the electronics, including the electric propellors, were sourced though AliExpress in China. Each propeller motor have their own power driver that connect directly to the autonomous pilot, that has one of the computers used in the project. This controls the boats steering and provides the ability to hold position when required. There are two Raspberry Pi’s also in the boat handling the interface and communication for both the weather instruments and the autonomous pilot. Open Source software ARDUPILOT has been used to send the course and instructions to the the boat from a remote location, which last Saturday was in the start box. The course is set using Google maps and the application allows a marine chart overlay.

Caelan and Alistair preparing to launch EV Spring

The power system consists of two 320 Watt Hour Lithium batteries, chosen to minimise weight and because they were the biggest battery that could be installed. Last Saturday at WBBC the boat travelled a bit over 6 kilometres and used a little under 10% of the battery capacity.

Communication has been the biggest problem in the project. Initially the design used radio, and the modems purchased claimed a 40km line of sight range. But although trialling various antenna the best range attainable was a bit over 300m. The communication failure caused near disaster with the boat attempting to escape or attempting to commit suicide on rocks. The Rhib we used for testing had to do some quick rescues. The poor communication nearly killed the project, and a couple of months ago the radio communication had to be replaced using the mobile phone cellular network. This has worked much better and the weather data is streamed to 2 computers handling communication and the database at our home in Pukerua Bay.

While a lot of the electronics, especially for the autonomous boat was outsourced, there was a lot of interfaces and software that had to be created to enable everything to be interconnected and work. This included the database and the dashboard that displays the weather information. In regard to NCEA and NZ Scholarship the project is finished. Some enhancements are planned once my other NCEA exams are completed. This includes raising the height of the weather instruments, adding air and water temperature sensors, and reducing the number of computers currently being used in the project.

Check out EV Spring on trials at Plimmerton