The presentation was "Using Robotics as a Learning Tool" by Colleen van Lent, Assistant Professor of Computer Science at California State University Long Beach. She started out with some background on her earlier research. Her area was not robotics, but she was lucky enough to do her graduate work at the University of Pittsburgh that had a large, multi-disciplinary research project in collaboration with Carnegie Mellon University and later the University of Michigan. Roboticists, computer scientists, human interaction experts and health care professionals combined their efforts to develop a robotic assistant, called Nursebot, to help elderly people with mild dementia. With help from the other specialists the computer scientists developed software that could provide help with the person's ADL's (Activities of Daily Living) such as bathing, eating, sleeping, medication and social activities.

The elderly population is growing at a huge rate and there are not enough caregivers available. Robotic assistants can provide planning and reminding people to do daily activities without being intrusive or commanding. They can provide electronic communication such as email, web services or teleconferencing without freaking a person out. Other alternatives have their disadvantages. The technology is difficult to transfer from one "smart house" to another. Palm pilot like E-Elves devices are easily lost or misplaced by the forgetful. Besides, robots are cool. Professor van Lent showed some photographs of elderly people looking at and using the Nursebots. A lot of time and money was spent developing an user friendly device that was attractive and non-threatening. The Nursebot developers did much interactive testing with elderly people so they could get feedback to improve the design.

Eventually she completed her degree and moved into academia where she discovered the truth that teaching is hard. How do you engage student interest without exceeding the limits of what can or should be expected from students learning basic concepts? The goal is to create a classroom environment that helps demonstrate how simple concepts can be used to tackle large-scale problems. There was also the culture shock of moving from a resource rich graduate environment on a major research project to being a low resource teaching professor. The learning environment is vastly different. Student motivations vary. Many students are just investigating the topic for the first time and are unsure of their interest level. Some students are openly hostile to the topic and others are frightened by it. The teaching responsibility is to provide a 50 minute lecture followed by a 75 minute lab. There are no TA's (Teaching Assistants) and each professor is busy so there is little cooperation. Money is much less available and she had to store the equipment in her office.

The real goal was to teach programming and problem solving to a diverse group of students, many of whom were working over 20 hours a week. Some were taking an initial course in their major and should be interested. Others were not majors, but had related science and engineering majors where computer knowledge was needed to pursue their subjects. Then there were the real nonmajors, those who might be taking the course because an introductory course in computing was required, but that didn’t seem to the student to be directly related to what they really wanted to learn. They had varying degrees of computer expertise and some were just getting started. Teaching problem solving is very hard. Open-ended problems seldom get finished, are difficult to grade and make it difficult to give feedback to students. Overly specified problems are boring and quite detached from the real world. It is hard to develop interesting problems within the scope of an introductory course.

Her solution is to use low cost robots to intrigue students new to the field, encourage teamwork and challenge students with previous experience. Using robots allows presenting problems in an open-ended fashion, provides the opportunity for creative programming, and allows the teacher to embed simple applications into interesting "cool" laboratory experiments. The equipment had to cost less than $10,000.00, the equipment had to fit into the teacher's office, and the equipment units must be capable of being transported independently. The complexity of the robot was hidden as the class was about programming and problem solving, not building robots.

There were some sample robots available at this ACM talk, as well as pictures and descriptions of them. She explained how some of the robotic movements demonstrated the use of various computer concepts. One example was a robot with two motors and wheels, a servo motor, a light sensor, and a sonar. This robot could be programmed for various activities from simply traveling back and forth, traveling in a circle, racing from a starting point to a finish line and back, following something at a distance, and finding the closest object. The students could directly observe how changing variables changed how the robots worked.

There were a number of problems because low cost is not always high quality. Repairs could often be obtained by feeding pizza to volunteers. The cost of pizza was a non-reimbursed cost to the teacher but was the quickest and most economical way of getting things done. There were IRB (Institution Review Board) issues about the adequacy of unique training procedures and the lab tests provided quantitatively sketchy results with regard to improved learning. The sensors were noisy, but was that really bad? It provided a "real world" problem that had to be resolved by the students. The program was a success as the students loved the robots. Students came to extra labs and participated in outside activities.

Professor Van Lent described additional robotic activities including outreach programs. There is a Lego League competition, Robo-Cup Junior and a Botball program. Palos Verdes high school even entered the major DARPA Grand Challenge program which requires a robotic vehicle to travel hundreds of miles under its own automatic control. The competitors included major universities and industrial companies and 40 qualified for the event. Palos Verdes didn't win the competition but was one of the 7 teams that actually started the race. No team won as the DARPA contest is currently well beyond the state of the art demonstrated by current robots.

Botball is run by the KIPR [KISS(Keep It Simple Stupid) Institute for Practical Robotics]. There is a 6 week program that begins with teacher instruction and ends in a regional competition. There are common tasks to identify team balls, cups and other things and transport them to the goal area and the teams try to remove opponent's objects. The schools receive free tutoring and keep all equipment for the relatively low cost of about $2,000.00. Teachers can demonstrate theoretical concepts (circumference, torque, gear ratios). Some organizations sponsor minority and female teams. Volunteer mentors can provide advice on feasibility, code debugging, web design and provide moral support to the students. Sponsors can interact either directly with local schools or through KIPR.

Professor van Lent provided an interesting description of a unique and innovative program to teach programming and problem solving using robots. Most elementary computing courses spend their time on small programs that don't really do anything but demonstrate mathematical concepts. Experienced programmers may know that these are related to real world problems, but this is not usually apparent to beginning students. Actually seeing your work either succeed in driving a robot or failing to do so properly should be interesting to almost everyone. It was also interesting for our audience to see the robots she brought to the meeting demonstrate their capabilities.