Dell develops mood-reading tech [310]

Jai Menon told the BBC that Dell Research was working on software for existing brain activity headsets that would accurately identify a wearer's emotional states.

He added that he believed such tech had uses in both the office and home.

But some experts are sceptical about the project

"I think the potential for these things is astronomical, but we've been told this technology has been five years away for decades," said Dr Bernie Hogan, a human-computer interaction expert from the University of Oxford.

Mr Menon, however, is more optimistic.

He said a team of two researchers were working with headsets made by Neurosky and other manufacturers - which cost between £60 to £120 - to see if they could be used to give a reliable indication of whether the wearer was happy, sad, bored or frustrated.

"If I can sense the user is working hard on a task, an intuitive computer system might then reduce distractions, such as allowing incoming phone calls to go directly to voicemail and not letting the user be disturbed," he suggested.

"Similarly, if they've been concentrating [for] a long time, maybe it could suggest a break."

He added that the kit could also be adapted for gamers - a market Dell already targets with its Alienware PCs.

"If someone is playing a game and it senses they are bored, it could ratchet up the level of challenge automatically. If it senses they are frustrated, maybe it's time to offer them a clue about how to proceed."

Dell is not the only major tech company to investigate the idea.

Microsoft has announced a series of mood-reading research projectss, including Moodscope - software to infer a user's mental state from their smartphone use - and a "smart bra" that monitors heart and skin activity to detect stress and emotions.

IBM has tested uses for brain-monitoring gear at its research base in Hursley, England.

And the video games company Valve also experimented with building sensors into a controller that would adapt gameplay to the player's body state, but later abandoned the idea.

Mr Menon said the current version of the software being tested was only able to correctly identify a headset wearer's mood about half of the time, but added he expected the figure would improve.

"We're trying to push the accuracy of our software into the 90% or better range, and if we can get there then the product starts to make sense.

"If an individual device doesn't give us that accuracy then we will also add additional inputs - a pulse oximeter [to monitor the level of oxygen in a patient's blood] or ECG (electrocardiogram - a heart rhythm monitor) or other readings, to see if multiple inputs help the software get to the correct value."

Brain-monitoring hardware is already used to let people with severe disabilities control computers and wheelchairs with thought.

In addition, a London-based start-up recently paired a Neurosky headset with Google Glass to let wearers use their brainwaves to make the eyewear take photos.

But one academic suggested that identifying moods would prove more difficult.

"Taking a picture is a fairly simple thing, it's like an on-off switch," said Dr Simone Stumpf from the Centre for Human-Computer Interaction at City University London.

"With moods or emotional states, you have a range of states.

"The headsets are fairly intrusive as well, especially if you want ones with lots of reliable inputs.

"And with some of them to get a good connection you need to use saline solutionution on the pads - imagine wearing that for 12 hours, it's not reallyly feasible."

Even if a mood-reading product did prove effective, Dr Hogan suggested many workers would prove resistant.

"The rights to my internal mind-state would be up for grabs," he said.

"Will it be a condition of my job that I wear something that monitors my mood? That's extremely scary and very different from a brain-control interface, which is more compelling."

Mr Menon stressed that Dell Research was only trialling mood-reading tech at this stage, and not designinging a product for market.

But he added that he still believed his firm could release a software solutionution soon.

"My goal is to work on interesting things and then persuade the rest of the company to build the products," he said.

"But I suspect that within a three-year timeframe, if the experiments are successful, then such products can certainly be available."