Monal Chokshi (D-Cog/HCI Lab, UCSD)

Utilizing drivers’ conceptual models of critical driving situations to aid the design of an Intelligent Driver Support System
The overall goal of the Nissan project is the design of an Intelligent Driver Support System, a vehicle-based technological system that acts as an extension of the driver’s cognition, especially during potentially critical situations. This system currently assesses the criticality of situations via engineering algorithms. However, these calculations of criticality have been shown to differ substantially from drivers’ notions of what comprises criticality.

How are drivers constructing their accounts of their concerns and of critical situations in their driving experiences? I am investigating this question for my second year project research. With this analysis, I aim to demonstrate how the integration of the driver’s perspective is vital to a successful design of the system. In my talk, I will present my approach to the analysis of the data, as well as show some examples from the data.

Amaya Becvar (D-Cog/HCI Lab, UCSD)

Video-Based Reflective Tools for Developing Expertise in Hands-on Learning Environments
Understanding what comprises expertise, and how it develops, has long been a major question to those studying human cognition. I am interested in the learning processes that go on in hands-on domains as people develop expertise, and how to design new instructional technology that might aid in learning the skills needed to become a competent member of a profession. In order to develop technology to teach expertise, it will be important to explore the learning effects of certain kinds of representations and instructional strategies.

I am engaged in the beginning stages of a 2 year ethnographic research project at a local professional education program in dental hygiene looking at how students develop expertise in the hands-on clinical laboratory portion of the program. During the fall and winter of 2004-5, I have been using ethnographic methods in order to analyze what counts as 'expertise' in dental hygiene instrumentation, and how these skills are communicated and developed in students during hands-on clinical practice. A major preliminary finding of the research has been that what counts as expertise to community members is often a problematic negotiation amongst faculty and students in the program. Expertise is currently neither consistently distributed amongst faculty and uniformly assessed in students.

In the subsequent phases of the project (2005-2006+), I plan to study how the use of new video-based reflective learning technology impacts how expert practices are coordinated amongst faculty and students, how students learn dental hygiene instrumentation skills, and how expertise is developed and assessed by students (and faculty) in the program. Video-based technology has great potential for making expert practices transparent to community members, and improving the standardization of expert practices. In this talk, I will discuss preliminary findings using video examples from in-class exercises and interview transcripts. I will also discuss some of the tentative technology impementation projects I am thinking to pursue in future work.

Ed Hutchins (D-Cog/HCI Lab, UCSD)

Discussion: Ethnographers' Digital Notebook
Ed will lead a discussion of the following publication, which reflects ideas of this lab. Please read this short, 4-page article before attending the meeting.

Field Notes on Field Notes: Informing Technology Support for Biologists
http://hci.stanford.edu/research/biology/publications/cstr2004-04.pdf

Neil McCurdy (Dept of Computer Science, UCSD)

We are rapidly moving toward a world where personal networked video cameras are ubiquitous. Already, camera-equipped cell phones are becoming commonplace. Imagine being able to tap into all of these live video feeds to remotely explore the world in real-time. We introduce RealityFlythrough, a telepresence system that makes this vision possible. By situating live 2d video feeds in a 3d model of the world, RealityFlythrough allows any space to be explored remotely. No special cameras, tripods, rigs, scaffolding, or lighting is required to create the model, and no lengthy preprocessing of images is necessary. Rather than try to achieve photorealism at every point in space, we instead focus on providing the user with a sense of how the video streams relate to one another spatially. By providing cues in the form of dynamic transitions, we can approximate photorealistic telepresence while harnessing cameras "in the wild." This talk will describe the RealityFlythrough system, and report on a live flythrough experience. We find that telepresence can work in the wild using only commodity hardware and off-the-shelf software, and that imperfect transitions are sensible and provide a compelling user experience.

Nissan-IDSS Project Team (D-Cog/HCI Lab, UCSD)

Ethnographic contribution to Intelligent Driver Support System (IDSS) design: guardians of the driver's perspective in a multi-disciplinary research effort
The Nissan IDSS three-year research project is coming to a close. The final workshop was canceled at the last moment, so instead of forcing you to pretend you are Nissan engineers (by giving you a practice talk), this will be a more Dcog-friendly discussion of our efforts to maintain an ethnographic attitude in the midst of a very experimentally oriented research network. We believe this experience left a significant impact on both the participating experimentalists and the ethnographers.

The biggest challenge to the ethnographers in doing research for safety system design (in contrast to style design) is the requirement of generalized findings across a large pool of subjects (and huge amounts of data!). Here the typical ethnographic recipe, which selects 'interesting' moments to do a detailed case-study on, is not sufficient. The crucial information from the engineering design side is about quantifying variability (between drivers and between situations) in the data. Our challenge was to find ways of describing situations and driver behavior in ways that map onto how subjects make meaning of their driving experience, on the one hand, and (descriptions) that would still be amenable to computational modeling of the data, on the other.

We will show how we developed an iterative process in which ethnographic immersion in the data (the usual arduous hand-coding) was combined with the development of automatic data processing. These efforts come together in a visualization tool that allows dynamic interaction with the multitude of video, audio, and other sensor data-streams from an instrumented vehicle. We are settling on using Petri Nets to bring together the ethnographic coding on one hand the mining of the data on the other. The topics we focused on in our research were partly determined by the research network (read, Nissan engineering concerns about rear end collisions), and partially arose from our stance as participant observers. Thus we will briefly touch on “Lane Changing”, “Car Following”, Relational Dynamics among drivers, and the opportunity to add Heart Rate measurement to the mix.

Dan Bauer (D-Cog/HCI Lab, UCSD)

The Semi-Coherent Practice of Design Grounded in Observation
I'll crudely barnstorm through some of the D-Cog themes we know and love: Simon's "Science of Design", activity landscapes, "representational talkback" with material anchors, and the co-evolution of designs and the behaviors they induce. I'll shoehorn it all into a tidy Theory of Everything, depicted by cute little boxes and arrows.

Tim Sohn (Dept of Computer Science, UCSD)

Wide deployment of location-aware applications is becoming more feasible due to the proliferation of location sensing mechanisms. Cell phones provide a compelling platform for location-aware applications, due to their large deployment and usability while on the move. This talk describes Place Lab, a radio beacon-based approach to location that can overcome the lack of ubiquity and high-cost found in existing location sensing approaches. Place Lab supports client-only location sensing on the cell phone using both GSM and Bluetooth "beacons" in the environment. Community-driven "war-driving" is used to build a database of beacon coordinates using GPS. Place Lab allows researchers to quickly develop and deploy location-aware research applications; one recent example being a social location disclosure application allowing users to manually or automatically share their location with others, or request others' locations.

Beth Simon (Dept of Computer Science, University of San Diego)

Classroom Lecturing, Classroom Learning: Technology to Engage a Diverse Student Body
The university classroom is not a static environment. The percent of people enrolled in post-secondary education grew 35% from 1980 to 1997, spurred by the earning potential afforded by further education. Did this growth merely lead to a larger number of the same "types" of students on campuses? Or has the increase in college-bound students increased the diversity of types of learners in our classrooms? I introduce Classroom Presenter, a Tablet PC based lecturing environment that allows for the exploration of more diverse modes of communication and learning styles in the classroom. Specifically, I will highlight the Student Submissions component of Presenter, which leverages technology to facilitate active learning or formative assessment activities in class. Students, equipped with Tablet PCs, can work on brief activities designed to reinforce or assess their grasp of a concept then wirelessly and anonymously submit their results to the instructor machine. Instructors can review student work privately to evaluate comprehension and project selected submissions to the class to highlight the variety of solutions or expose common misconceptions. Additionally, I'll describe Ubiquitous Presenter – a system designed to bring similar affordances to students with non-Tablet internet enabled devices. Finally, I'll briefly outline some of my other ongoing projects in Computer Science education involving technology, exploration of student understanding of design, and novice code reading and tracing skills.

Project Updates & Open Discussion Forum