Weed Strategy Simulator for Farmers
xCog is a new data simulator to support farmers to decide on the best weeding strategies based on their farming options, possible weather conditions and aligned to their personal priorities. Great data simulations rely heavily on not only good data, but on well-defined logic that processes the presentation of data according to real world scenarios. For me to design both the system (data formatting and logic for processing and feedback) and user interface (mechanics, and aesthetics for user control and observation) comes with intense collaboration with experts in the related fields being represented in the system. In this case Kate Devitt (cognitive) and Tristan Perez (decision scientist) provided the insights to best design and validate the data logic. These experts are also encouraged to work directly with my team – programmer, Ryan Quagliata and 3D artist, Wade Taylor – to tweak all aspects of the simulation to ensure it’s relevancy and accuracy.
The system uses graphical representations of the components of the decision problem and allows the user to consider different scenarios by adjusting their priorities and other contingencies. This is then combined with an animated representation of the scenarios that can help trigger positive cognitive experiences and subsequently, application in the real world that it represents.
This project is an example of the interactive modeling systems can be designed for people to experiment with the consequences of decisions within certain systems and ecologies. These modelling systems are often presented as playful interfaces but have smart underlying systems and complex algorithms to present data, situations, choices, people, places and things in a way that is interactive and attractive but also challenging and transformative.
Team
Deb Polson: Research and Design
Ryan Quagliata: Programmer
Wade Taylor: 3D Artist
Lead Researchers
Kate Devitt (Cognitive Scientist, QUT), Tristan Perez (Decision Scientist, QUT)
David Thornby (BYGUM Weed Expert)
Funding Body
Institute for Future Environments, QUT ($43,7500)
Data Simulation
This work is part of a folio of data simulation works starting with SCAPE 2005, FarmIt Right 2008, ECOS 2013 and Future Fruit 2016. I explicitly refer to this body of work as data simulations rather than the simpler data visualisations or infographics… as I intend to do more than present pre-processed insights and instead offer data controls in real scenarios that are simulated to respond to users’ experiments with complex systems. These controls connect multiple data sets within any given scenario with logic rules and a computational system that processes complex decisions in real time within the interface. More significant as it is a stronger motivator of change as users can experiment with the consequences by exploring options and challenging pre-exisiting expectations and strategies. This combination of visual and statistical data agency and feedback can effect peoples action and promote radical change. As a sharable device, the simulations can elevate group discussions well beyond present and derivative details as teams can collaboratively experiment and discuss impacts and risks of decisions that effect the future of economies, environments and cultures.
