The Ocean Hackathon® is an annual competitive event which involves cities across the globe to compete to solve several ocean-based challenges using multiple disciplines of science together by young innovators. The challenges involved disciplines such as computer science using machine learning, AI, web, and app development & marine science including biology, oceanography, atmospheric science. The Ocean Hackathon® is a 48-hour event of decoding the oceans that takes place simultaneously across different cities worldwide. Cape Town was the first African city to be part of this global event in 2021 and has stayed on since, encouraging other African countries to join the global initiative. Cape Town was the only representative of Africa this year at the Ocean Hackathon®.
The 2023 Ocean Hackathon® took place at the Le Wagon, in Cape Town and it came with 4 new challenges for young innovators to tackle. It started on 17 November 2023, 17h00 to 19 November 2023, 17h00, making it 48 hours! Four different teams with young innovators from different institutions such as UCT, CPUT, Stellenbosch, UWC, & We Think Code had members with different skillsets to be able to tackle the challenges given. The four challenges presented to the teams were as follows:
- Challenge 1: Citizen engagement & eco-action for marine ecosystems – with GreenWay Africa (Pty) Ltd. – CITIZEN CAM
- Challenge 2: Predict the sardine run with image magic – with SAAMBR – The South African Association for Marine Biological Research. - SARDASH
- Challenge 3: Dive into Kelp Farm Tycoon: Explore “Re-wilding the Oceans” – with Kelp Blue. - AGRIKELP
- Challenge 4: Code Eyes for our Aerial Ocean Guardians – with Seafar. – GUARDIAN AI
The teams came up with names (in bold above) to bring that competition spirit aboard. We had data coaches throughout the hackathon to help us acquire and understand the data we needed and made use of.
My team came up with the name SARDASH (Sardine + Dash) as we took on the second challenge which required us to create a digital tool that uses historical data of environmental parameters and algorithms to predict sardine movement and their predators. The team consisted of five members, myself - Andile Kuzwayo (UCT), Fatsani Phiri (UCT), Nomvula Mpungose (UCT), Thobeka Sithole (We Think Code) & Xolani Mvana (We Think Code), and our data coach was Mark Addison (Blue Wilderness Dive Expeditions). It took a lot of work to try and understand how to go about this challenge but we eventually started working with machine learning and historical data from sources such as KZN Sharks Board, SAWS and NOAA. Throughout those 48 hours we were able to develop a prototype web application that could simulate a potential path of sardines and named the site, CHASERS.
On the final day (19th Nov) of the hackathon at 17h00, we had to present our prototypes and solutions to the given challenges in front of a panel of judges of different disciplines and organisations such as City of Cape Town, Two Oceans Aquarium, Ocean Hub Africa, marine and computer scientists. After the presentations were done and the judges deliberated, they picked team GUARDIAN AI as the runner -up and excitingly my team, SARDASH came first with team cash prizes of R12 000 and R22 000 respectively. The winning team also gets the chance to go to Brest, France to compete against other teams that won the competition in their respective cities for the number one spot across the world. The team will also get to develop their prototype further so that it can be implemented for worldwide use.