A team of University of Sydney students gained second spot against teams of researchers in the AgData Challenge hackathon, with a PhD student also leading the winning CSIRO team.
A team of eight PhD students led and mentored by Agriculture and Environment’s Associate Professor Tom Bishop have taken out second prize for the AgData Challenge, awarded today at the CSIRO’s premier agriculture and food event, AgCatalyst.
The event attracted more than 550 representatives from business, peak industry bodies, producer groups, research funders and providers and policy-makers.
It reminds us that our own PhD research should consider the end use and the end user
The University of Sydney team presented their methodology, results and user-friendly models – all products of a three month agricultural data hacking competition run by the CSIRO with data provided by their industry partner, Lawson Grains.
The intense and highly competitive hackathon involved creating a value-added product for a corporate farm.
Ten teams were involved in the competition, eight of which were from CSIRO, one team from Queensland University of Technology and the sole entry from the University of Sydney.
The three best teams participating in the Challenge competed for the top spot yesterday. The prizes were awarded alongside a panel discussion about how to extract more value from agricultural data, with industry and research leaders Chris Sounness from Birchip Cropping Group, Leanne Wiseman from Griffith University, Jen Taylor from CSIRO, and Adrian Turner from Data61.
Associate Professor Tom Bishop said the team of PhD students was ecstatic to gain second place in a crowd including computer scientists, statisticians, agricultural scientists and economists.
“We are also incredibly proud to say that Ross Searle, who led the winning CSIRO team is also completing his PhD at the University of Sydney’s Faculty of Agriculture and Environment whilst working at the CSIRO,” Associate Professor Bishop said.
Patrick Filippi, one of the eight Agriculture and Environment PhD students who participated in the data hackathon, said it was motivating to know that the models they created and the information from the data could have real-life, practical applications and uses for growers to improve their yield, sustainability and profitability.
The team used the provided data to produce a model that could predict crop yield for the coming season.
They took a precision agriculture approach, suggesting variable nitrogen fertiliser rates within a paddock based off the predicted crop yield. This could be done at two time points; sowing – when an initial nitrogen application takes place, and at top-dressing – when the final nitrogen application occurs.
“The models we produced were of a high quality, could detect fine resolution spatial variation well, and could be used to make many more management and economic decisions on top of our variable nitrogen fertiliser suggestions,” Mr Filippi said.
“We placed all of our models in a user-friendly online interface where growers could easily use the models and see the results for their own crops.
“We have valued the opportunity to be involved in the hackathon; this kind of task is transferrable to a real-life situation and has practical implications for farmers and data analysts. We can see its usefulness and it reminds us that our own PhD research should consider the end use and the end user, and that our research will help to improve the sustainability of growing food.”
Data hacking will become a vital component of our future work, and will enable us to make the most value from agricultural data
Dean of the Faculty of Agriculture and Environment and Professor of Soil Science, Professor Alex McBratney said today’s result was testament to the forward thinking and innovative approaches that the University encourages its students to undertake.
“Data hacking will become a vital component of our future work, and will enable us to make the most value from agricultural data,” Professor McBratney said.
“For 25 years we have been leading the way with our work in precision agriculture at the University of Sydney, working to improve profitability and sustainability on-farm by pinpointing the requirements of broad acre farms at a fine spatial resolution.
“As we move into a new era of developments in technology and big data analysis, we have the background and scope to expand our horizons in research, analysis and modelling, to work collaboratively and take the lead in the emergence of digitally de-commoditised agriculture.
“In the new year, I will take on the new and additional role of Professor of Digital Agriculture as we move the University of Sydney into the forefront of digital value-adding,” he added.
“Our focus on developing digital agriculture will further increase on-farm profitability and sustainability and will enhance quality, differentiation and traceability of agricultural products to further boost consumer confidence which is so vitally important as our products make their way around the world and we concrete Australia’s reputation for high quality, safe foods,” said the Dean.
“I’d like to extend my congratulations to Tom and the team, and Ross and his team on these wonderful achievements.
“The outlook is incredibly encouraging and I look forward to the contributions of our young researchers to the future wave of agriculture,” Professor McBratney said.