SYDNEY – April 16, 2019 – Six AI-focused Australian projects have been awarded support through Microsoft’s $US50 million AI for Earth program. Australian recipients this year include Monash University; Griffith University; Queensland University of Technology; InFarm; the Australian Wildlife Conservancy; and, Bush Heritage Australia.
AI for Earth is designed to harness technology to help mitigate and adapt to changing climates, ensure resilient water supplies, sustainably feed a population rapidly growing to 10 billion people, and stem the ongoing and catastrophic loss of biodiversity.
The need for rapid action in Australia has been reinforced by recent environmental challenges including extended drought, dry-lightning triggered bushfires in Tasmania and Victoria, record high temperatures across the nation, Townsville’s flood, and the devastation wrought to agriculture across the Top End by a combination of flooding rains and overnight low temperatures.
AI for Earth is more than just grants – Microsoft is helping to bring transformative solutions to commercial scale and offering open source API solutions to help organisations everywhere scale their impact.
The program’s participants receive a mix of Microsoft Azure cloud computing resources (including AI tools) and/or data labelling services, as well as access to training on Microsoft’s innovative data science, machine learning, and visualisation tools to support the most scalable, innovative solutions in the areas of climate change, agriculture, biodiversity and water – so we are no longer flying blind to the needs and challenges facing our planet.
According to Lucas Joppa, Chief Environmental Officer, Microsoft; “As a technology company with a deep commitment to sustainability, we understand that our responsibility extends beyond our own operations to innovating towards a healthier and better future more broadly.
“The world is seeing rapid advancements in cloud and AI solutions that are unlocking new possibilities to solve the world’s most challenging problems. But the uptake of those solutions to understand and protect the planet is proceeding slowly, and as such, we are essentially flying blind when it comes to understanding how our planet is changing and how to best solve environmental challenges. AI can change that.
“Time is too short and current human resources are too few to solve urgent climate related challenges without the exponential power of AI. By putting AI in the hands of researchers and organisations we can use important data insights to help solve issues related to water, agriculture, biodiversity and climate change.”
The Australian projects which have this year received AI for Earth grants and support are:
Mapping species distributions: This Faculty of Information Technology project at Monash University uses social network geotagged photos and harnesses Azure cognitive services to provide insight on key ecological phenomena, including insect pollinator distributions and flowering plant activity, to understand how they are impacted by climate change. According to PhD candidate Moataz Medhat ElQadi and Associate Professor Alan Dorin; “Access to Microsoft’s Azure Cognitive Services is enabling us to continue to build an understanding of ecosystems key to food security and the sustainability of earth’s natural environments.”
Below ground carbon level prediction: Using Azure machine learning, Griffith University is developing a prediction model providing greater insights regarding carbon sequestration of green stormwater infrastructure. Integrating different levels of relevant ecological data with state-of-the-art AI methods to facilitate the better prediction of the urban carbon budget. PhD student Emad Kavehei noted; “Traditionally, we monitor a few influencing factors of soil carbon accumulation while we know that there are more factors involved.” Integrating different levels of relevant data with the state-of-the-art AI methods better enables the team to predict the urban carbon budget.
Drone-based reef monitoring: A Queensland University of Technology project exploring how data collected by drones fitted with advanced sensors can be interpreted by AI to aid with the classification and restoration of coral. One of the challenges is the scale of the reef – which is roughly the size of Japan, requiring an innovative approach to data collection. Associate Professor Felipe Gonzalez explained that the team has built an AI system that interprets images captured by hyperspectral cameras mounted on drones, to get a better understanding of the health of the reef.
Weed identification and classification: Startup business InFarm is using AI to help identify weed species in fallow fields – specifically those that are chemical resistant – and also to provide application maps for use in autonomous, variable spray tractors. Jerome Leray, managing director of InFarm said: “We use drones to survey fields, AI to interpret the images, and the resulting insight to allow targeted weed-spraying that we estimate can reduce farmers’ herbicide costs by 95 per cent and also rein in the amount of chemicals used overall.”
Invasive predator detection in the outback: Using 60,000 camera trap images, leading conservation organisation Bush Heritage Australia and data scientists Jonathan Bourne and Anindya Basu (ODINN) are using Azure virtual machines to develop a species recognition algorithm to estimate the predator population on their properties, which can help optimise predator detection and protection of threatened species.
Feral and native fauna identification: Using AI and image recognition, the Australian Wildlife Conservancy is exploring how to automate animal identification from 90,000 camera trap images.
Microsoft’s belief in AI-enabled impact extends beyond AI for Earth which is part of the company’s broader AI for Good initiative.
Together AI for Earth, AI for Accessibility and AI for Humanitarian Action represent a $US115 million five-year commitment to combine Microsoft’s technology and expertise in AI and data science with the talent and expertise of groups around the world in fields such as environmental science, disability needs and humanitarian assistance.
These latest six Australian projects are among the 230 spread across more than 60 countries which have received AI for Earth grants and support to date.
Since the program’s inception researchers are reporting benefits from faster processing speeds, improved accuracy in algorithms, and the ability to scale through new APIs.
This helps boost our ability to address and adapt to a changing climate promising a more informed, sustainable and resilient future.