A couple of weeks ago our ECE department hosted Bret Harsham of Mitsubishi Electric Research Labs (MERL) as a part of the ECE900 graduate seminar. Bret gave a very interesting lecture about the ways of shortening voice dialogs through using a contextual push-to-talk button. The title of the lecture was “Contextual Push-to-Talk: Shortening Voice Dialogs to Improve Driving Performance.”

The focus of the lecture was on a prototype in-car voice user interface (VUI) which was tested during my MERL internship last summer. As opposed to the contemporary in-car VUIs, which use only one push-to-talk button for issuing commands, this work presents a way of utilizing multiple push-to-talk buttons depending on the context of the query. For example, if we have three domains of interest, we can associate one push to talk button for each of them. Therefore, we skip multiple steps which are otherwise required in order to switch to a desired domain and initiate a voice search.

This work was recently accepted for publication at MobileHCI 2010 conference.

Zeljko Medenica

Last week my colleague Zeljko Medenica and I visited AgeLab of the Massachussetts Institute of Technology. The visit was organized by the New England Chapter of the Human Factors and Ergonomics Society (NEC-HFES). This laboratory is investigating the role of technology in the human aging process. One of the big issues that older people face is driving safety. Bryan Reimer gave a very interesting lecture on driving research done at the lab.

He emphasized that in order to know more about people’s driving behavior, it is very important to know their workload. The workload measures they use are very similar to the ones we collect in our experiments: driving performance, physiological measures and subjective ratings.
After the lecture, we had a hands-on experience with the MIT AgeLab test vehicle and driving simulator, thanks to Bryan, Jarrod Orszulak and Bruce Mehler. The proprietary Volvo test car allows them to collect naturalistic driving data using many kinds of sensors. The driving simulator is based on an STI simulation engine. The car cab is a fixed-base VW Beetle placed in front of a projection screen.

In one of the studies done at AgeLab, researchers compared results from the driving simulator with naturalistic data and found that physiological measures correlate very well over the different platforms. It was very interesting to hear the experience of fellow workload estimation researchers and compare methodologies and measures used. Thanks Brian, Jarrod and Bruce for hosting us.

Oszkar Palinko

Continue Reading »

This summer I have had a great opportunity to be an intern at Mitsubishi Electric Research Labs (MERL) in Cambridge, MA. This is not only a great prospect for my professional career, but also a chance to experience how working in a real company looks like.

MERL is a daughter company of the Mitsubishi Electric Corporation from Japan and therefore they have close collaboration. As the name implies, it is a research facility where people work on many different areas, such as digital communications, multimedia, user interfaces, speech interaction, mechatronics and many others. Their overwhelming publications page just confirms how important research is in this institution.

As a MERL intern, I am a member of the speech group. My advisor here is Bret Harsham and for this summer we have been working on testing an in-car speech user interface which was developed here at MERL. This interface enables contacts, music and points-of-interest selection using voice commands. The experiments will be concerned with the influence this interface may have on driving performance and will be performed on their driving simulator (shown in the picture below).

The driving simulator is based on a racing game simulation and consists of three huge DLP projector screens which create a very wide field-of-view, force feedback steering wheel and pedals, and a motion chair. The motion chair is very powerful, because it simulates the vibrations caused by road surface and engine, as well as the tilting of the car caused by acceleration and deceleration. The feeling it produces is very realistic and it may help prevent simulator sickness which was so common in our simulator studies.

We are looking forward to publishing the results of this interesting study. So, stay tuned for more info on this topic.

Zeljko Medenica

Hello ecebloggers,

Over the past semester Andrew Kun, Tim Paek, Zeljko Medenica, Oskar Palinko, and I have been investigating the influence of Personal Navigation Devices (PNDs) on driving. Nowadays these devices are widespread in vehicles. They typically display the real-time location of the vehicle on a map and play spoken prompts when drivers need to turn. While such devices are less distracting than paper directions, their graphical display may distract users from their primary task of driving.

We have conducted experiments in our high fidelity driving simulator to see the influence of two PNDs on driving:

• the influence of a standard PND, with a combined visual (electronic map) and spoken output,
  
• and the influence of a PND with spoken output only.

The first goal of the experiments was to find out if a PND with combined output causes drivers to spend less time looking on the road ahead than a PND with spoken output only. The second goal was to examine the effects of glancing at the PND’s visual display on driving performance. We hypothesized that a PND with combined output will influence the percent dwell time (PDT) on the outside world negatively when compared to the PND with spoken output only, i.e. drivers will be more distracted by it. Our second hypothesis was that PND’s visual display will influence negatively on the driving performance. We expected an increase in the lane position and steering wheel angle variances to occur whenever drivers look away from the road.

The experiment results showed that our hypotheses were correct:

• We found that drivers using a navigation system with a graphical display indeed spent less time looking at the road compared to those using a navigation system with spoken directions only.
• Glancing at the display was correlated with higher variance in driving performance measures.

The video below shows how the PND’s visual display influences driving performance.

With respect to designing in-car navigation aids, our results seem to suggest that if users can trust a PND enough to follow the spoken directions they are given, even when they are lost, a navigation system with no visual display may be the most favorable option since visual attention and consequently driving performance will likely be improved.

This work will be presented at the AutomtoiveUI2009 conference.

Have a good one,

Nemanja Memarovic

On Thursday, April 2nd 2009, Paul Green visited our lab. At that time he gave a talk on Human Factors Engineering and spent the whole day at our lab talking to students and visitors of the Open House.

Dr Green is a distinguished Professor at UMTRI and the president of the Human Factors and Ergonomics Society. He has immense experience in human factors engineering for vehicles. He was kind to share this experience with Project54 students and staff.

In his lecture Dr Green emphasized that vehicle accidents are one of the leading causes of preventable fatalities in the US. Therefore it makes perfect sense to research how human factors engineering can provide a safer in-car environment. Many passive safety systems are already deployed in cars (safety belts, airbags, child seats, etc.) Active systems are also finding their way into commercially available vehicles (anti-lock braking system, electronic stability control, adaptive cruise control, etc.) A so-called Workload Manager is a cutting-edge active safety system. Dr Green presented us his research on how to design such a system. What can an everyday driver take home from this research: it is a bad idea to have any additional tasks while driving, especially when maneuvering (turn, overtake, lane change, etc.) The talk was very well visited by UNH students, faculty and others.

After the talk Dr Green led an informal discussion panel at the Open House event, where everyone could try our driving simulator. We had many visitors from different UNH offices as well as from other institutions.

During the day, we had several opportunities to have discussions with Dr Green. Andrew, Alex, Zeljko and Nemanja were very interested to hear Paul’s experience with running experiments as well as about his opinion on our research plans. Dr Green emphasized many times during these talks that, to produce “good research” it is necessary to identify and precisely define the problems that need to be addressed. Furthermore, hypotheses are even better if they are accompanied by quantitative predictions. Also, he stressed the importance of repeatability of scientific research in our field: it is important to mention all important factors when writing a paper and have the same measures of performance throughout the research/development community.

From left: Andrew, Oszkar, Alex, Paul, Nemanja and Zeljko

All-in-all, Dr Green’s visit was a very inspiring one. I hope that we all will be even more able and motivated to do even better research thanks to Paul.

Oszkar Palinko

Last Saturday, March 28th, the University of New Hampshire had its annual College of Engineering and Physical Science (CEPS) Open House for incoming Freshmen. This event is designed to share the work and research that is being done at the university by the various faculty and staff with prospective students and their families. Through this event, CATLab - Project54 was able to share our recent work and research that has been done with the driving simulator. Students were able to experience the simulator first hand, and see the work that goes into the research that we do.

Pictured here we see Oskar Palinko showing the simulator to a few students and their parents, giving them all the opportunity to try it out for themselves.

Here Ed Bourbeau talks about his time working for Project54 and the many skills that have served him well while working in industry as an electrical engineer.

Here I am showing a student, and his younger brother, the driving simulator. Who knows, he maybe a future student?

Overall the event went great! Those of us here at Project54 felt honored to have been able to show the next generation of engineers our work, hoping that some of them may have been inspired to pursue a interest in something they may have seen. There were many excellent questions asked, and a fun time was had by all! Hope to see the new Freshmen in the Fall Semester!

Matthew Lape

Last Wednesday our Project54 lab was one of the hosts at the High Technology Day here at UNH. Lots of local high school students had a chance to learn first-hand about different areas that various laboratories at UNH’s electrical, computer engineering and computer science departments work on. Oskar, Mark, Erika and me had a chance to introduce students to the work that we do in our lab.
Mark gave presentations about our Project54 system and also showed how the system works when installed in a real police cruiser. The picture below shows Mark demonstrating the system in our “lab car”.

Oskar and I presented our driving simulator and some of the interesting research we perform on it. Almost always when we present our lab, the simulator gets the most attention and this time was no exception. Each group of students had a chance to experience the simulator by driving in the virtual world. As expected, most students couldn’t resist trying what happens in the case of a collision, which is exactly why simulators are preferred in this kind of research. Picture below shows how the demonstration looked like.

More information about this UNH event can be found on the Foster’s Daily Democrat website. I would also like to thank Erika for taking great pictures of this event.

Zeljko Medenica

Check out the video below in which Ivan Tashev, Mike Seltzer and Y.C. Ju discuss the Commute UX project at TechFest 2009. 

Last summer Oszkar and I visited the MSR driving simulator lab and it’s great to see how quickly Ivan and co. are making progress with the simulator. And it’s pretty cool that the video features ABBA!

UPDATE: The video above doesn’t feature the MSR driving simulator, but in the video below Ivan Tashev introduces Commute UX using the simulator - thanks for the comment Oszkar.

Andrew Kun

On April 2nd, 2009 our lab will host Paul Green, a distinguished Research Professor at UMTRI and the President of the Human Factors and Ergonomics Society. Dr. Green will give a lecture on Human Factors Engineering as a part of the CEPS Frontiers Lecture Series.
The lecture will also be an integral part of the ECE900 Seminar Course. Dr. Green has many years of experience in conducting research using driving simulators. He has more than 200 publications and he was the lead author of SAE recommended practices concerning in-car navigation systems design.

After the lecture, Project54 will host a Driving Simulator Open House, during which Dr. Green will be available for questions and everyone will have a chance to try our high fidelity driving simulator.

Also, visitors will be able to get informed about new Project54 applications, updates, experiments. The schedule for April 2nd, will be as follows:

12:30 Light lunch at Kingsbury Hall N101, Durham, NH
1:00 Lecture: Human Factors Engineering by Paul Green at Kingsbury N101
2:00 Driving Simulator Open House with Paul Green at Morse Hall 229 (refreshments and snack)

Everyone is welcome! Come, visit us and learn more about ergonomics and Project54!

Oszkar