We have four confirmed projects for MISG 2016, and are working to get one or two more.

Inference in a knowledgebase (DST Group)

Knowledge can be represented by a graph of entities (e.g. people, locations, events) and links between entities (e.g. “is related to”, “visited”). The attributes of entities and the attributes of links are uncertain, and the probabilities associated with entity attributes and link attributes are generally not independent. How can uncertain attributes be represented? How can inferences generated from the knowledgebase be ranked by certainty?

Sequencing ore extraction to control blend quality (Schneider Electric)

Mine scheduling is the task of determining the best material to supply to a processing plant in order to hit production targets, subject to a number of physical, logical, and capacity constraints. This includes determining an excavation sequence for excavators to follow and selecting material to draw from stockpiles to service the processing plant. On extraction, material can either be processed directly or stockpiled for processing at a later time. A geological model specifies quality attributes of material in the ground, but this model has a level of uncertainty associated with it. The problem is to optimise the extraction sequence and processing plant feed in order to maximise the chance of hitting the production quality and tonnage targets. The planning method must be scalable to work with tens of thousands of distinct “blocks” of material in the geological model.

Modelling water pollutant density associated with surface water runoff (SA Water)

South Australian Water Corporation sources a significant portion of its raw water from surface water catchments and the River Murray. The raw water is then treated before delivery to customers. Elevated stream flow resulting from rain events is a major driver behind the transport of land-based pollutants to surface water. Ideally, raw water quality should be measured at times of peak flow, but this is not alway possible. How can we optimise our sampling to capture a range of runoff events, and then predict pollutant density under a broad range of conditions including peak flow?

Optimisation of household PV and storage (Ergon Energy)

Rooftop solar photovoltaic (PV) panels, household electrical energy storage (batteries), home energy management, interval metering and new tariffs will change the way that households use electricity from the grid. Distributed storage can also give electricity retailers the ability to shift loads in response to changes in the wholesale price of electricity and constraints on the distribution network. What is the ideal mix of PV, storage and tariff for a customer? What is the value of these technologies to customers and to electricity retailers?