Celebrate Fresh Thinking: Professorial Lecture Series

Join us in celebrating the very substantive contribution to academia made by Professor Richard Green and Professor Adrian McDonald in the first Professorial Lecture Series for 2019.

  • Date: Thursday 4 April, from 4.30 – 6.00 p.m.
  • Location: E14 – Engineering Core

All staff and postgraduate students are encouraged to attend the lecture series to actively support our new Professors, and take the opportunity to appreciate the fantastic research being undertaken in parts of the University you may be less familiar with.

You’ll find further information on each presentation, below.

Ngā mihi

Professor Ian Wright
Deputy Vice-Chancellor | Tumu Tuarua

 

Presentation details:

Deep learning is Revolutionising Smart Robots, Drones and Vehicles
Presented by Professor Richard Green, Department of Computer Science & Software Engineering.

Over only the last few years, deep learning has significantly improved computer vision and machine learning to finally enable fully autonomous robots, drones, vehicles and other analysis usually performed manually.

Helping computers to unambiguously see and understand the world is a fascinating and exciting research endeavour, but also seriously challenging when they are still so dumb, and almost blind compared with human cognitive capability.

In this Professorial Lecture I will describe my contributions across these research areas, including recent autonomous systems research into drones pruning forests, robots pruning vineyards, general purpose farm robots, autonomous underwater vehicles (AUVs) inspecting mussel lines, scanning ship hulls and wharf pylons to detect invasive bio-fouling species, mapping the seabed to locate scallops and automatic blood spatter analysis.

UC is now a world leader in this autonomous software research, with a large AI Robotics (UC AIR) research group. But this is multidisciplinary research – which only exists through collaboration with Electrical, Mechanical and Civil Engineering, together with domain experts such as those from marine biology, forestry, ESR and Lincoln University Viticulture.

Labour shortages and quality/accuracy/safety are the biggest drivers for so many of these applications – but based on this accelerated capability, how long will it be until all human manual labour tasks can be automated?

 

Clouds over Antarctica and the Southern Ocean: Their Influence on Aotearoa New Zealand Climate
Presented by Professor Adrian McDonald, Department of Physical & Chemical Sciences.

Clouds have a surprisingly large effect on our climate. In particular, cloud cover reflects sunlight back into space that would otherwise be absorbed by oceans, potentially raising their temperatures.

Despite their significant influence on climate, clouds still represent the largest source of uncertainty in modern climate models.  The frequency of clouds over the Southern Ocean for example is often underestimated, causing models to predict warmer sea surface temperatures than what is observed. 

Models also often misrepresent the composition of clouds because of the importance of small particles, called aerosols, which act as the starting points for cloud droplets and ice to form around. 

These deficiencies in turn leads models to predict the strength and position of the Southern Hemisphere storm tracks incorrectly. These storm tracks impact Aotearoa New Zealand directly via their influence on rainfall, and also bring extreme weather events. It is vital that our models represent clouds well so we can increase certainty in our climate projections for Aotearoa New Zealand.

In this Professorial Lecture, I discuss the use of detailed measurements to compare with simulations of the present-day to critically test the quality of cloud and aerosol simulations.

By analysing these differences and using our understanding of how cloud processes work I show how we can develop improved model simulations.

2 thoughts on “Celebrate Fresh Thinking: Professorial Lecture Series”

Leave a Reply

Your email address will not be published. Required fields are marked *