Identifying a gap between our theoretical understanding of wireless senso-actuator network technology and its real-world applications, Associate Professor Dani

Cause for pause: Hopping from hardwires to no wires at all

Identifying a gap between our theoretical understanding of wireless senso-actuator network technology and its real-world applications, Associate Professor Daniel Quevedo is seeking to remedy shortfalls of time-honoured paradigms in engineering science and bridge 'what is' with 'what could be'.

Associate Professor Daniel Quevedo

Associate Professor Daniel Quevedo is a problem solver. Combining elements of systems control, signal processing, and wireless communications to develop original, high-performance engineering designs, he is also scientifically curious about everyday technology.

'My research is about finding ways to enhance the existing uses of embedded processors for the control of networked systems,' he explains.

'I want to enable novel functionality too.'

Building on the emergence of networked systems, as well as what we already know about associated wireless and computational technology, Quevedo is looking to provide solid theoretical foundations and practical solutions for real-world problems. Similarly addressing the current challenges these new technologies present, the Chilean native is duly working to revise contemporary understanding of industrial control processes and cyberphysical systems.

'It is by no means clear how to harness their potential,' he admits.

'A common thread is that many of the standard paradigms which allow the separation of computation, communications, and control are no longer valid.'

Quevedo began addressing such inadequacies when he moved to Australia to undertake a PhD in 2001.

Opting to study a specific class of engineering design problem where decision variables are limited to a prescribed set of values, his approach complemented elements of systems control with those of signal processing and information theory. Quevedo also sought to design model predictive controllers for power electronics during his candidateship, working within available constraints to improve future and current energy conversion methods. In doing so, he opened the door to access a variety of situations where traditional methods have only been able to give partial and unsatisfactory answers.

'I was able to formulate and resolve design problems within a solid quantised control framework,' he declares.

'This allows for powerful solutions to be obtained in a systematic manner.'

From stumbling blocks to stepping stones

Chiefly interested in understanding how we can take full advantage of networked control system technology in factory settings, the engineering scientist is looking at ways to replace outdated hardwired communication links.

'The advantages of wireless are enormous,' he enthuses.

'It's much easier to install and maintain, it saves space and weight, and it increases system flexibility.'

'Wireless devices can also be placed where wires cannot go, as well as where power sockets are not available.'

To design new closed loop control systems so that they work over the current protocols for wireless communications, Quevedo is studying measurements of wireless sensors set up in factories and other buildings. He is also in the process of building a mathematical model of the situation and investigating the types of optimisations needed to achieve efficient operation.

'Our key challenge is to overcome the bottlenecks caused by limitations in communication and computational resources,' he states.

'But we still need to respect fundamental stability and performance limitations.'

International industry engagement

Quevedo has also had a hand in research that led to two international patents.

The first saw him collaborate with Ericsson to develop power control and algorithms for cellular systems. Working on the basic condition that increased power usage leads to decreased battery life, the team analysed mobile phone activity at Ericsson's base stations to better understand network tradeoffs and their implications for the average user. Also wanting to avoid associated problems with system overload, they formulated a precise balance between sustained cellular reception and power usage.

'To make efficient use of available resources, these controls are essential,' Quevedo says.

The trilinguist's more recent patent similarly underlies many industrial applications. This time collaborating with ABB Corporate Research Switzerland, a global leader in automation and power technologies, Quevedo modified existing control designs to boost the performance of power converters in renewable energy and smart grid contexts.

Well-travelled and well-equipped

Juggling publications in top international journals, external thesis examinations for universities in Germany and France, as well as expert assessments for the Australian Research Council and Chilean Commission of Scientific Technological Research, Quevedo is soon set to throw a few more responsibilities into the mix. Having just been named runner-up in the Engineering and Technology Category for the 2014 Scopus Young Researcher of the Year Awards, this list of accomplishments is vast and ongoing.

Perhaps most impressively, next year he will give a semi-plenary talk at the International Federation of Automatic Control Conference on Nonlinear Model Predictive Control. Quevedo is also Editor of the International Journal of Robust and Nonlinear Control, Associate Editor of the Australian Control Conference, and was a member of PhD Candidacy Examination Committees at Notre Dame University in the United States and Stellenbosch University in South Africa.

Long-term logistics

Believing networked control systems should 'simply work,' even in the presence of uncertainties in the availability of energy, computation and communication resources, Quevedo anticipates a winding road ahead for engineering science and related technology fields.

'It will all come down to how well we're able to translate the plug-and-play functionality of the Internet to these distributed systems,' he says.

'We'll need to learn from established design techniques and develop new ones.'