Closing out 2023, changes to Aotearoa New Zealand’s climate were more apparent than ever, with major weather events shaking the country, such as the Northland flash flooding and Cyclone Gabrielle. Such events show the importance of using data to make better predictions, and the importance of working together to learn from these to move forward.
A whakatauki (Māori proverb) comes to mind: Kia whakatōmuri te haere whakamua – I walk backwards into the future with my eyes fixed on my past. It reminds us to learn from history while embracing change.
We celebrated a big milestone last year, we are now in entering the second half of the TAIAO programme! Below is a recap of our achievements for 2023, we are looking forward to many more in 2024 and beyond.
We are excited to embark on our fourth year in 2024, which will be enriched by the valuable input and feedback we received during last year's TAIAO workshop. It was truly rewarding to connect with members of our community, igniting fresh opportunities for collaboration and research.
For those who could not join us either in person or virtually, recordings of the workshop are available on our YouTube Channel (@TAIAO_AI). Simply navigate to the TAIAO Workshop 2023 playlist and select any talk that piques your interest.
Hackers came together over two days last August to build exciting projects and create a collaborative environment targeted at solving a specific problem. Throughout the hackathon, teams learnt many new skills, met like-minded peers, and networked with sponsors and mentors.
Participants were encouraged to bring their own problems to solve, with 24 teams taking part. The winner was “Deep Weed - AI for proactive and scalable management of weeds”.
Four teams took their ideas to the Aotearoa AI Summit in September, we can not wait to see their ideas develop further in 2024!
We were very proud to partner with Indigidata Aotearoa this year, a programme open to Māori (tauira, kaitiaki, kaipakihi) who want to better understand the opportunities and challenges associated with Māori data science, Māori data sovereignty, as well as key technical, cultural and ethical issues.
TAIAO presented to 20 participants on machine learning and coding, you can find more information on the programme here.
Our Changing Coasts is a five-year MBIE funded programme bringing together world leading researchers to help forecast sea levels over the next 100 years, and how our coastal landscapes will change. This programme, led by Professor Richard Levy from Victoria University of Wellington, builds on NZ SeaRise, which produced sea level projections to include state-of-the-art global climate data and estimates of local vertical land movements around Aotearoa New Zealand.
The team have partnered with TAIAO and will be using machine learning and detailed process-based models to better understand where our most vulnerable coastlines are. Read more here.
In early 2023, AI and Data Science Professor Yun Sing Koh, research fellow Ben Halstead and their team of researchers were awarded for their research on Probabilistic Framework (PF) methods at the Data Science and Advanced Analytics conference.
Read more about their research and the team's new PF 'SELeCT' here.
In 2023, we continued to develop a new AI system to forecast the impact of future flooding events, particularly in the Coromandel region. This system uses AI and machine learning to predict floods by analysing multiple data streams, including rainfall measures, river water levels, and rain gauge catchment data. The goal is to enhance real-time predictive capabilities, with a prototype expected in the Coromandel, and potential further roll-out across New Zealand.
Aotearoa Species Classifier
Our comprehensive species classifier for New Zealand's unique flora and fauna has had some exciting enhancements last year. Utilising over 1.2 million expert-annotated images, the AI program can identify a wide range of species, providing scientific and common names, among other classifications. A publicly available app launched earlier in the year which even works without an internet connection and a more powerful web classifier offers advanced models and visualisations.
Algal Blooms Prediction
A machine learning algorithm has been developed to estimate the quantity and type of algae in New Zealand lakes using satellite imagery. This approach aims to provide a more representative and cost-effective method than traditional lab analyses. The machine learning model is trained with a small set of data points, combined with unlabelled data, to build accurate predictive models. The research has potential applications for monitoring waterway health by companies and councils and has received recognition at an international AI conference.
We would like to take this opportunity to say thanks to all our partners, collaborators, and community that support TAIAO. Happy New Year and we look forward to continuing to work together to see more great outcomes in 2024.
TAIAO provides an open source community platform for Environmental Data Science and AI. In the platform, you can discuss the notebooks and datasets found on TAIAO and get to know other contributors. Click here to sign up.
Subscribe to our newsletter! Click here.