Both photos: ©2000 Artville, LLC. A Strategy for Integrating Robot Design Competitions into Courses in Order to Maximize Learning Experience and Promote Intellectual Development esign competitions are becoming increasingly com- design activities [4]. One outcome of this emphasis has been mon, especially in the field of mobile robots. the incorporation of national design competitions di- The sponsorship of such competitions rectly into the classwork of its “Introduction to ranges from local clubs of enthusi- Robotics and Computer Vision”(junior/se- asts to large professional organi- nior level), “Advanced Robotics and Com- Dzations, such as the American Association puter Vision” (graduate level), and “Field for Artificial Intelligence (AAAI), which Session” (computer science design sponsors the annual AAAI Mobile Robot practicum) courses offered by the De- Competition and Exhibition as part of partment of Mathematical and Com- their annual conference. Awards for com- puter Sciences. The intent of these petitions vary from certificates of merit to efforts is to improve students’ design significant prize money; for example, the competence and intellectual maturity by Association for Unmanned Vehicle Sys- taking advantage of the opportunities af- tems (AUVS) Annual Ground Robotics forded by robot competitions. Competition awards $5,000 to its first-place This article reviews experiences associ- winner, $3,000 to the second place, and $2,000 to ated with these courses and their design teams the third place. over a five-year period. The design teams met the The Colorado School of Mines (CSM) has been a leader in objectives of each course while earning one first-place and fostering the intellectual development of its students through four third-place awards plus $5,000 in prizes as well as produc- 44 IEEE Robotics & Automation Magazine 1070-9932/01/$10.00©2001IEEE JUNE 2001 ing seven papers. Based on these positive experiences, I en- N to evaluate progress of the community in some aspect of courage others to integrate competitions into classes and offer mobile robotics (for example, walking platforms); recommendations on how to facilitate the process. Although N to encourage undergraduate students to get more expe- the focus of this article is specific to classroom education, it is rience with mobile robotics, thereby improving both expected that these observations will also provide insight into the quality of students going into the workforce and the the incorporation of design competitions into the general ed- number of students willing to make a career in robotics; ucational experience. A competition can reinforce key con- N to expose undergraduate students to faculty and research cepts of a course and bring reality to an AI robotics project, issues, thereby encouraging them to consider graduate while producing a viable entry. Other teaching opportunities school; not covered in class, such a budgeting, improving communi- N to encourage graduate students and their advisors to cation skills, etc., are not discussed in this article. consider pressing but unsolved issues facing robotics, in effect, “mini” grand challenges of a field; The Perry Model suggests that a N to encourage interdisciplinary interactions competition can aid the intellectual between academic researchers in the hopes that such endeavors will become common- maturity of students who are beginning to place; and accept that there may be more than one N to have fun. The potential impact of these competitions on right answer to a problem. research should not be underestimated. The AAAI national conference hosts increasing numbers of The article begins by contrasting the objectives of competi- demonstrations of software agents playing competitive games. tions versus educational experiences. The conclusion is that certain portions of competitions can be an important tool for Competitions and the Perry Model fostering intellectual maturity, as defined by the Perry Model While competitions clearly offer something to the research [3]. Next, a classification of the major international robot com- community, they also offer a more immediate payoff to the petitions is presented. Background information on the courses general educational development of the individual student. A is presented next as well as how that influenced matching classes competition provides additional extrinsic motivation for the to a specific competition. How the courses were organized to students to mature according to the Perry model of intellectual take advantage of the competition topics is discussed next, with development [1]. Perry’s model defines nine stages of increas- particular emphasis on grading and class size, the spectrum of ing complex reasoning. Paraphrasing, positions 1 and 2, dual- philosophies for managing competitions, and the difficulties ism, reflect a student’s attitude that right and wrong answers with cooperative learning. A section on fielding a team by inte- exist for all problems. At position 3, early multiplicity, students grating the class projects is also included. The article concludes realize that knowledge includes methods for solving problems with a synthesis of the experiences at CSM into a general strat- and that there may be more than one right answer to a prob- egy for incorporating a competition into a robotics course. lem. Students reach position 4, late multiplicity, when they be- gin to think and analyze about the diversity of possible solutions. At position 5, relativism, students evaluate solutions Competition versus Education from different contexts. At positions 6-9, commitment within rel- The goals of a competition often have a surface intersection ativism, the students are able to take into account that the with the topics addressed by education. However, competi- world is a changing place. tions focus on winning (although some define “winning” in A competition involves a clearly defined but open-ended terms of design rather than performance), whereas education problem to which there are many possible solutions. The stu- concentrates on teaching the methods that ultimately lead to dents must be largely self-reliant to understand the problem success. Instructors use a variety of techniques in the class- and to apply their knowledge and problem-solving methods; room, most notably laboratory exercises and projects, to en- they must function at the least at the early multiplicity position. sure that each student has the opportunity to meet the learning Working voluntarily with other students encourages the stu- objectives. The Perry Model suggests that a competition can dent to identify and evaluate a variety of opinions stemming aid the intellectual maturity of students who are beginning to from the scientific literature, the professors, and other stu- accept that there may be more than one right answer to a dents, which should move the student to a late multiplicity posi- problem. The primary issue for an instructor is how to identify and integrate the appropriate aspects of a competition into the tion. The typical team organization of an advisor plus student laboratory and project assignments. members lends itself to a realization of relativism, whereby the teacher serves as a consultant instead of an arbiter. The goals of Competition Goals the competition, as opposed to other potential applications of The sponsors of robot competitions generally cite the follow- robotics, serve to stress the contextual aspect of applying ing reasons for the competition: knowledge. JUNE 2001 IEEE Robotics & Automation Magazine 45 In addition to the general intellectual maturation of the stu- dent’s tendency to shun teamwork (“I could have done it dent, a competition provides a forum in which to acquire and better by myself.”) or to transfer responsibility to another exercise specific job-related skills. Mobile robot competitions, agent (“It was my partner’s/the hardware’s/the soft- in particular, require sophisticated software, interdisciplinary ware’s/the competition sponsor’s fault). The competition interactions, and teamwork. It is difficult for the students to experience should be orchestrated so that the students have a make progress without applying good software engineering reasonable chance to succeed if they proceed correctly, to principles, especially testing and debugging strategies. The discern the cause of failures as they occur, and to make cor- rections if possible to prevent such situations from occurring in future projects. The rela- The competition experience should be tively low level of intellectual maturity of orchestrated so that the students have a undergraduates favors the instructor provid- ing the overall architecture and design so reasonable chance to succeed if they proceed that the students work on portions that are correctly, to discern the cause of failures as both relevant to the course and have a high potential for being completed. they occur, and to make corrections if Shelter students from extraneous concerns. possible to prevent such situations from Competitions often involve many activities occurring in future projects. that would not be encountered in a normal laboratory experience: acquiring/maintain- ing hardware, raising money, etc. While students will probably need to collaborate with students from these activities are likely to be part of a student’s future career, other disciplines; for example, the smooth control of a mobile they can take valuable time and attention away from the ob- robot depends both on the power train and