Wired Mag interviews me on Eyetracking

Posted on 22/03/2006 by admin

It’s been an incredible (and tough I’d say) starting of week: I had a lot of works at home to make it a little bit more secure against intrusions (and I’m not speaking about hackers ).

But I also have been so lucky to have a long chat with Frank Rose, contributing editor at Wired Magazine about the Eye Tracking research projects I’m currenlty working on, stressing those on remote control for Television/Video content: they’re both based on a remote usage of eyetracker in order to interact with the content on the video. In one of those the user can unconsciously rotate the real 3d scene selecting the character/object she preferes.

My opinion is that existing TV remote controls are unusable when the user needs a deeper interaction with the filmed scenes since she has to look down at the remote to find the buttons that need to be presses, while – using eye tracking technology – she could easily watch at the screen and select what she likes/needs without looking somewhere else.

It was a deep 40mins chat that I won’t be able to summarize here, I’ll just wait for Frank’s piece being published.

Innovative uses of Eyetracking technology

Posted on 14/03/2006 by admin

UPDATE late 2009: the idea behind aDAMS has been successfully patented with number ITMI20050494

I’ve nearly finished to prepare the slides for my talk at the Innovation Expo in Milan this Thursday 16th (more details on time/location here): I’ll introduce aDAMS (anti-Drowsiness Alert Management System), an eye-tracker based system which continuously scans the driver’s eyes to evaluate her attention leve and properly activate a scalable alert.

After receiving the data we elaborate a series of analyses based on position of the eyes (mapping on the correspondent street target), size of the pupil, frequency of saccades, blinking activity, etc. etc.; the system also has networking capabilities in order to propagate the alarm. Being base on eye tracking means we’re using infrared light, thus being independent from light condition and being the HCI interaction absolutely natural, with no side effects to the driver.

I know there’s a bunch of other anti-drowsiness systems based on the use of similar technologies, but I pretty sure our systems outstands all the competitor.

Interested in? Curious? Drop me an email or come for a chat on Thursday.

Clicktracking VS Eyetracking

Posted on 10/03/2006 by admin

It’s a while since I’ve noticed a Web 2.0 application called Crazy Egg which will show you (for free!) were the users clicked on your page using eye tracking analysis derived graphics such as heatmaps and some brand new visualizations for the overlays (colophon: my employer sells eyetracking services).

Let’s take a deep breath and try to understand whether or not this amount of data could be useful “to improve the effectiveness of your web site”.

I’d say that if you’re a part-time blogger or run a small ezine this kind of data could be of some utility to you, but if you’re a serious web application developer/designer or want to test the usability of your company /client site interface well, this kind or data are more or less garbage.

I strongly disagree with what is being said by Nick in the Weareseencreative post on the subject : you should really care what your users are looking at, otherwise you won’t be able to understand (and thus fix) why they aren’t clicking on a particular link or button (BTW Nick, the Tobii eyetracking suite we’re use tracks also the users’ clicks).

Moreover the users’ interactions with a site/interfaces cannot and shouldn’t be summarized to just the clicks, there a lot more: the images, the texts and the page designs which actively participate in designing the user experience.
The click is just the final step of a longer decision path thus showing where the user clicked definitely doesn’t explain WHY she clicked: I think that the most significative quote I can place here is that what really impressed me when I first approached eyetracking technology is that the mouse pointer is completely STOPPED during the whole decisional process; out of sight, I’d say).
And I think that is really important to designers and developers to understand WHAT the user looked at before clicking and WHY they looked at or interacted in a given way (and this is simply impossible with a mousetracking machine).

I don’t want to be misunderstood: my opinion is that Crazy Egg could really help to improve the overall usability of small and simple sites showing their creators what is clicked in a graphical way (a more complexe alternative could be a deeper stats/logs analysis, and my friend Lou could say more on this) but once compared to Eyetracking technology, its benefits simply disappear.