IT tuning Aussies into emotions

Monitoring traffic emotionally

Meanwhile, researchers at the Sydney-based NICTA labs have been working on a similar concept for the NSW Traffic Management Centre called STaR UI. Research group manager and senior principal researcher, Fang Chen, said its aim was to create a more user-friendly interface for traffic operators taking care of the state's traffic.

"At the same time we try to understand their mental effort or so called 'cognitive load' involved in their ongoing daily basis work; how they can cope with that, if there are any other strategies to help them do their work more efficiently or effectively," Chen said.

Currently at the prototype stage, the technology utilises speech recognition and video and image processors to evaluate and combine different modalities - like speech, hand movement, body movements and some typical mouse/keyboard movements - and then provide a meaningful response.

"The information can be captured by a hand movement," Chen said. "For example, there are a lot of police cameras in different places but if an accident happens and you really want to locate a camera to look at you can point to that one and it will switch to that camera's view."

Other than pioneering technological utility, the system also aims to help manage stress loads."Because we capture all these modalities' information we actually use this information to get an estimation on a user's mental load," Chen said. "Like when you are suffering some stress or high mental load your speech actually gets changed [sic]. So we try to measure all this and make an adaptive interface based on these measurements."

In future, Chen hopes the technology will help the vision and hearing impaired and also be used in industries that have a centralised control room with a complex system and information flow, like bushfi re management, air traffic control, mining, power plants and defence.

"In future you will be able to use your natural modality to deal with machines in your everyday normal life," she claimed.

The challenges

Due to the enormous variety in human behaviour and idiosyncrasies, one of the biggest challenges both Khosla and Chen face is setting parameters for evaluation.

"That's a huge challenge, you have to set some sort of benchmark," Chen said. "Let the computer know what kind of changes that means, if it is a significant change for a human or just a little bit of a change."

While not specifically challenging the work of Khosla and Chen, other researchers caution that evaluating emotions or mental states through technology can be misleading.

University of Technology Sydney researcher, Toni Robertson, who supervises several technology-based Phd projects and worked on a system called Bystander that exhibited crime photos at the Museum of Justice in NSW, questioned our ability to accurately formalise parameters and evaluate social reactions.

"We know if it is something that can be measured it can be misused," Robertson said. "To me the way you would humanise technology is actually make it easier for people to use it. To me this means making it more like technology and designing it very well so what it does is good."

But while the realm of emotional profiling is still in its infancy, the innovation driven by these kinds of technologies has the potential to not only showcase pioneering Australian IT, but also help us understand more about our emotions at work. Or at least remind us to do some exercise and not have that second helping at lunch.