We develop smart machines, robots and visual sensors to perform day to day operations such as harvesting and grading based on companies’ needs in horticultural and agricultural fields. We work in collaboration with partners such as Robotics Plus, ArborGen, Tendertips and Boyds in developing solutions
to problems mainly in (but not limited to) the outdoor environment. Some example projects are the MaraaTech MBIE project, researching robotic vine pruning, apple fruitlet thinning and blueberry harvesting. Autonomous Multipurpose Mobile Platform (AMMP) - an all-purpose driverless horticultural vehicle,
robotic kiwi fruit harvester, apple harvester and asparagus harvester.
Applied Machine Learning
While our group mostly focuses on developing solutions for industry to solve real-world problems in a commercial context, we also bring our expertise in software engineering and machine learning to the table to deliver working prototypes for research projects. Our general-purpose, open-source ADAMS
workflow engine is used commercially for processing spectral data, like NIR, MIR and XRF, in environmental laboratories (Ravensdown/ARL in NZ, Eurofins Agro in NL). ADAMS is also used in a research project with Plant and Food Research for analysing NIR spectra of kiwifruit. For image processing tasks,
such as image classification, image segmentation, object detection, we deliver deep-learning based solutions (Redfern Solutions, ESR, Zespri).
We have a long history developing time-of-flight range imaging sensors and are well known for proposing a number of solutions for mitigating errors in range imaging. Our collaborators include former Waikato time-of-flight people who successfully started the independent company Chronoptics. Currently
scene motion causes large errors in ranging, and we are working on transforming the problem of motion in the scene into velocity measurements. Other major projects include developing a biologically based visual sensor for mobile robotic navigation and research into the non-destructive evaluation of materials,
with structural health monitoring of components and structures using point sensors alongside imaging based approaches such as infrared thermography, thermoelastic stress analysis and digital image correlation.
- Royal Society Te Apārangi Early Career Research Excellence Award for Technology, Applied Science and Engineering - for developing
new techniques to improve a type of 3D imaging called time-of-flight - Lee Streeter.
- The KuDoS Datamars Engineering Science Award - Lee Streeter.
Our expertise is not limited to agricultural and horticultural robotics. One of our main goals is to apply robotic solutions to commercial problems and we can achieve this with our wide range of expertise as portrayed in other sub-groups: agrirobotics, applied machine learning and sensing. One example
project is the robotic asparagus harvester where we develop the vision sensing system that incorporates machine learning for detecting and locating harvestable spears, and the agrirobotics based end effectors to cut and pick the identified asparagus. Some of our completed projects also include cave auto-sampler,
lavender distiller system and also non-invasive neural interface based artificial hand.