Operations Research and Computational Analytics
Data-driven decision making informed by theory, algorithms, and computational methods to solve real-world problems.
The Operation Research and Computational Analytics (ORCA) group is a team of academics within the Department of Engineering Science who specialise in data-driven decision making. They utilise a variety of theory, algorithms, and computational methods to solve real-world problems with real-world data.
What is OR?
Operations Research (OR) is the science of utilising advanced mathematical theory/computational methods to help make better decisions. Also known as the “science of better”, OR utilises ideas from area including as mathematical modelling, optimisation, simulation statistical modelling, decisions analysis, and artificial intelligence to solve real-world problems.
What is CA?
Analytics is the capture, management, organisation, and visualisation of data to inform better decision making. Computational Analytics embeds this data into a mathematical/statistical model that enables the decisions making process to be automated and/or optimised. Computational Analytics is the science behind data-driven decision making in the real-world
Our stochastic programming research is motivated by practical problems in industry in which some optimal decision is sought subject to uncertainty on the problem data. These problems often originate from planning models in which the future is uncertain. Stochastic programming problems can often be formulated as equivalent mathematical programming problems with many variables and constraints, and techniques to solve them usually incorporate some form of decomposition. The development an improvement of these techniques is one focus of our research. Introductory material on stochastic programming can be found at the Stochastic Programming Society website.
Air crew scheduling
Air New Zealand has been using computer systems for crew scheduling based on OR research for many years. They use OR techniques developed by ORCA researchers to synchronize crew and aircraft routes to operate efficiently and decrease the probability of severe delays.
Telecommunications network design
We analysed and designed the coverage of fibre optic cables in the Auckland CBD. This network design was automatically generated using optimisation methods developed by ORCA researchers.
Optimising power generation and purchasing in electricity markets
Stochastic programming problems arise frequently in the electricity sector. Scheduling power generation and transmission has traditionally been a difficult optimization problem due to the uncertainty of supply of water to the hydro lakes, and the uncertainty of daily load. New models are currently under development for optimizing power generation in the wholesale electricity market, under the auspices of the Electric Power Optimisation Centre.
The ORCA group have had a long involvement over the years with the New Zealand yacht-racing fraternity. There is considerable interest amongst these groups in using optimization techniques in yacht design, performance analysis and weather modelling. Find details at the Yacht Research Unit.