Rationale By connecting the CS, LSIS, and BS students into creating the tools necessary to conduct field studies, we have accomplished an important goal of our undergraduate programs: building successful collaborative and integrative teams. The CS students learn the intricacies of software development for a real project with real stakes rather than a toy project. The LSIS students learn an understanding of the steps required to perform research with scientific rigor. Finally, the BS students have a valuable tool for scientific sampling while learning to apply statistical concepts to ecological communities. This one will go on the resume! - CS Software Engineering Student

Implementation (pedagogy) In the lecture-based classroom we are using the tablets on both sides of the classroom. The instructor uses the tablet connected to a projector in place of the blackboard. We are also using screen-recording software (Camtasia Studio) to record lectures and place them online as flash files. The students in the classroom use the tablets to take notes, easily create and brainstorm designs, and coordinate group work during breakout sessions. The computers allow individuals to perform their own work and easily share it with others using OneNote's shared sessions and other online tools. While the CS students are using the tablets to create tablet-enabled software, the BS and LSIS students are using the tablets in the field in combination with specialized devices (e.g. cameras, laser range finder). Rather than collect data using pencil and paper, data is collected directly in electronic format using the tablet and external digitization devices. This enables validation of the sampling data in real-time.

Moose Habitat Surveyor Screenshot Implementation (technology) Our model requires interaction between CS, BS, and LSIS students where the BS and LSIS students and faculty act as clients for the custom software developed by the CS students. To assist in the cross-disciplinary communication we have set up project wikis and blogs to inform everyone of progress, issues, and key developments. The CS students are using Java, Visual Studio, Visio, and project management tools to design, implement, and document the software. A variety of hardware devices have been integrated to create the field-based data collection application. The herbivore ecology application is integrating a GPS unit, Mitutoyo digimatic calipers, Laser Range Finder, and a Polhemus 3D digitizer to sample plant information for moose forage. The students are using HP TC4200 Tablet PC's to develop and collect data. We are also using a ruggedized Tablet PC, the Itronix DuoTouch, for fieldwork in rougher weather conditions. The LSIS students are using digital video cameras, radar guns, sound recorders, and 3rd party traffic modeling software on the Tablet PC's. For more detailed information and code downloads about the moose forage data collection system, see the link below. Moose Habitat Surveyor

Project Status Year 1 In this first year we focused on developing prototypes for the data collection software. Since the software needs to be somewhat mature and stable before the biology and natural science students could conduct field experiments, a majority of their participation was staggered to year two. However, the CS students interacted with biology and physics/astronomy faculty and students to implement the software. As a result, the CS students have learned valuable lessons in how research is conducted, how to develop tablet software, the importance of requirements, and the importance of involving the client in the software development process. One major challenge has been scheduling. Our original intent was for the natural science students to use the software in Spring of 2006 in collaboration with further development by the computer science students in the software engineering class. However, the software engineering class was offered in Fall of 2005, but canceled in Spring 2006 due to low enrollment. Additionally, the scope of the projects was too large to fully implement in one semester, which resulted in only partially working software for the Spring 2006 semester. As a result, we have had to adjust our schedule for collaboration to Fall 2006 and Spring 2007. At the end of the first year we had a working prototype of the Moose Habitat Surveyor and a partial prototype of the agent-based traffic simulator. Year 2 In the second year we unfortunately made little progress on the computer science side, but did have good progress on modeling traffic flow in the LSIS course. The software engineering course was offered in Fall 2006, but enrollment was low with an initial headcount of 11 students. The group working on the Moose Habitat Surveyor suffered greatly when a team member tragically died (in circumstances unrelated to the project). Shortly after, the team leader dropped the class. As a result the remaining two team members were unable to complete much work but we do have a functional prototype. The group working on the Traffic Simulation project suffered from incompatible team members. One of the team members did not get along with the rest of the team and in the end the entire team essentially dismissed the project portion of their grade and submitted a project that would not even compile. Unfortunately, this discord was not discovered until near the end of the semester when it was too late to reassign or make a significant intervention with the team members. Despite these failures, the CS students did learn valuable lessons in software engineering. Unfortunately, this did not help the LSIS students waiting for the software to be developed. In spite of the lack of agent-based traffic simulation software, the LSIS students made excellent progress collecting data and generating hypothesis on factors that influence traffic flow around the UAA campus. Armed with digital video cameras and radar guns, the students measured traffic densities and velocities at various intersections and crosswalks on campus. From this data they generated theories and predictions about what causes traffic congestion and proposed ways to improve it. Students were very receptive to the project and have remarked on how well they relate to it since everyone has experienced traffic congestion. As such, it has proven to be an excellent motivating project to teach students scientific methods and principles. Year 3 The computer science side of the project was on hold during year 3 as the CS faculty member was on sabbatical. However, further progress was made for the LSIS and Biology students. In the summer of 2007 our students modeled the roads around campus, using the tablet PCs and Google Earth images of campus to accurately trace out the road lengths and geometries. We also used the camcorders purchased with the grant to videotape intersections to determine the traffic flux densities in each direction, and to determine the timing of the lights. In the Fall students used these models to see how the intersection behaviour changes with different traffic and pedestrian loads during different times of the day, and to explore with the SimTraffic software what potential benefits there are to modifying the traffic flow (e.g., with timed pedestrian lights, extra lanes, restricted campus entrance/exit points, etc.) The herbivore ecology students used the prototype software and hardware to map and quantify the carrying capacity of habitats for moose of the Placer Valley on the Chugach National Forest. The students and Forest Service biologists spent several weekends in Placer Valley, and the students learned the techniques of vegetation surveying, GPS navigation, GIS classification of habitats (using ARC-GIS), and various nutritional techniques for forage analysis. The collaborative project was extremely successful, and culminated in a presentation to Supervisors, managers, and biologists of the Alaska Department of Fish and Game, the US Forest Service, and the University of Alaska Anchorage. Lessons Learned Attempts to extract field-usable software of a significant scale is very difficult to do using students in a software engineering class. This difficulty is compounded by other factors such as low enrollment and the motivation of the students. In the future the instructor plans to play a much more hands-on role in effectively serving as a group manager for each team, rather than the current hands-off approach. Additionally, a more agile software development methodology appears to be a better match for these projects than the traditional waterfall model, especially given the short time frame of a semester. Finally, there was not as much inter-disciplinary collaboration among students in the different disciplines as we hoped. Much of the difficulty was due to scheduling problems inherent with students enrolled in separate courses at separate times and in separate disciplines. In the future we intend to teach a cross-listed course where both CS students and Biology students can be enrolled together in the same class to work on projects. This model has been used successfully in the past with Bioinformatics courses. We have already completed our curriculum process and have created a new Honors course that can serve as a common elective for students from multiple disciplines to collaborate on a research project.

CS student Jim Weller testing the biomass data collection system The tablet and mobile capabilities of the equipment allows real-time data collection and hypothesis testing that was not possible just a few years ago. This project also builds upon the expertise of faculty and students in different disciplines. The collaborative arrangement makes the project feasible and beneficial for everyone involved. Moose Habitat Surveyor Video Short demo of the hardware, sampling in the field, and screen capture (WMV format).

Impact on Student Learning Three years ago there was very little collaboration between faculty and students in the computing, biological, and physical sciences despite many opportunities. We have made progress in fostering collaboration with the HP Technology for Teaching grant. Today, we now have CS students successfully collaborating with biology faculty and students to create software and implement models to solve real research problems. The Moose Habitat Surveyor system is near complete with a working prototype. Our assessment tools include student evaluation forms, concept mapping, laboratory assignments, interviews, in-class feedback, and monitoring of the usage, content, and degree of collaboration using our online communication tools. Assessment also consists of qualitative feedback regarding the efficacy of the software and the collaboration process. In the LSIS courses, students used the HP equipment to participate in a research project to study traffic flow around campus. We have collected pre and post concept maps that assess a student's understanding of ten concepts related to "scientific research" (e.g., data collection, analysis, hypothesis, results, and so on). Students showed an improvement in concept understanding, moving from an average of 2.7 pre test to 6.4 post. Action Steps from HP2006 Technology for Teaching Conference Measuring Impact Worksheet from HP2006 Technology for Teaching Conference

Impact on Teaching The goal of the software engineering course is to teach students how to develop medium to large-scale software beginning with requirements elicitation to testing, deployment, and quality assurance. The HP technology and collaboration with BS and LSIS provides an ideal environment to teach these concepts. The LSIS and BS courses focuses on teaching students how the process of science works by giving them the opportunity to participate in authentic scientific research supported by the HP equipment. The value of research-based learning is that it develops critical thinking and teamwork skills important in all fields of work. Furthermore, the technology allowed us to move outside of the classroom to collect data with the laptops and instruments. For example, students collected data by videotape, radar gun, sound levels, 3D digitizer, digital caliper, and laser-based distance meters around campus and in the field. While the traffic modeling project has been a huge hit with students, it has also allowed the instructor to model the scientific research environment and serve as a facilitator. In this role, students actively participate in the research process rather than listen to the instructor lecture about research processes. Because of the success of this project, we are continuing to receive funding from the US Forest Service for support of students on habitat research, and they will be providing additional support for a similar project in the Fall of 2008.

Quick Facts Departments: Mathematical Science, Biological Science, Liberal Studies Integrated Sciences Courses Impacted: CS A401- Software Engineering, CS A470 - Applied Software Development, LSIS A202 - Natural Science Concepts & Processes, BIOL A445 - Herbivore Ecology 70 students Impacted Three Faculty Involved This project is funded in part by an HP Technology for Teaching grant.

Kenrick Mock Contact Us PI: Kenrick Mock, Ph.D., Computer Science, kenrick@uaa.alaska.edu, (907) 786-1956 Co-PI's: Donald Spalinger, Ph.D., Biological Sciences, afdes@uaa.alaska.edu Travis Rector, Ph.D., Physics and Astronomy, aftar@uaa.alaska.edu

References & Publications Mock, K. (2004). Teaching with Tablet PC's. The Sixth Annual Consortium for Computing Sciences in Colleges Northwest Regional Conference, October 8-9, 2004, Salem, OR. Geopad Information Technology for Field Science Education and Research Moose/Deer Carrying Capacity Web Model Tablet PC Resources CS A401 Webpage Software Engineering Course Webpage Camtasia Studio Screen recording software Sample recorded lecture

This project supported in part by an HP Technology for Teaching grant.