Contact for general enquiries: Peter Baumgartner
As of February 2011. Interested EMCL students are advised to liaise with the given contact person.
Contact: Toby Walsh
The world is full of symmetry. For instance, we might have identical trucks to route, equivalently skilled people to roster. When solving combinatorial problems using search techniques like constraint programming, symmetry may be inherent in the problem or may arise through modelling it as a constraint problem (e.g. naming essentially indistinguishable objects). In either case, symmetry can lead to redundant search, as many symmetrically equivalent blind alleys may be explored wastefully. To avoid this, symmetry-breaking constraints can be added, to exclude all but one of each equivalence class of solutions. Alternatively search methods can be adapted to prevent them searching symmetric states to those already visited. The aim of this project is to develop new symmetry breaking methods.
Contact: Toby Walsh
A wide range of constraints can be specified and reasoned about using automata or formal languages. For example, the REGULAR constraint allows global constraints to be specified using a finite automaton or regular language. As a second example, the GRAMMAR constraint permits global constraints to be specified using a context-free grammar. We would like to explore other tools from formal language theory. For instance, can we use packrat parsers to specify and propagate constraints? As a second example, can we use a LR parser?
A second topic is to study encodings of global constraints into propositional satisfiability (SAT). A number of useful global constraints like REGULAR can be encoded into SAT. We can then to use a fast SAT solver to solve constraint satisfaction problems. The aim of this seconf project would be to explore other global constraints that can be encoded into SAT.
Contact: Toby Walsh
We are working closely with the Road Traffic Authority of New South Wales to develop new traffic controllers. The current focus of research is the control of multiple intersections and whole regions, as well as the opportunities that arise from new technologies like car to car communication. We are also working closely with a local start-up who are providing personalized travel planning services. There are many interesting research problems here. For instance, how do we capture user's travel preferences? How do we recommend travel routes that combine multiple modes of transport like bicycles and trains?
Contact: Toby Walsh
Companies like DHL and FedEx and many others routinely have large logistics problems to solve. They have parcels/engineers/lorries too that need to visit a number of customers/locations/supermarkets and they need to decide who visits whom in what order. The efficiency of their business depends on solving such problems quickly and efficiently. We are working closely with a number of industrial partners in developing the next generation of decision support software. There are many interesting research questions. For example, how do we make solutions robust to uncertainty (new orders arriving, traffic delays, etc)? How do we just past performance to improve future solutions?
Contact: Toby Walsh
Static analysis use automated tools to identify possible bugs in software. It can, for instance, identify problems with buffer sizes and pointer usage. It is especially useful in safety and security critical applications where correctness is important. We are interested in how the powerful reasoning tools from constraint programming can be applied to static analysis. For instance, does static analysis throw up new types of constraints for us to reason about? Can we use constraint solving to identify boundary cases?
The College of Engineering and Computer Science at the Australian National University runs a Summer Scholarship program. Most of the topics or related ones offered under this scheme might are also suitable for EMCL students. See here.
Disclaimer: The information given below is non-authoritative. The visa classes listed below are the ones that work from the NICTA perspective. Please see the immigration web pages for details, such as eligibility criteria, and how to apply.