REU Site: Collaborative Research: Developing, Analyzing, and Evaluating Self-drive Algorithms Using Real Street Legal Electric Vehicles on Campus

该项目为本科生提供了动手主动学习的机会，使用街道合法车辆进行城市道路自动驾驶功能的研究。美国本科生很少有机会使用真实车辆开发自动驾驶算法。在本研究中，学生将在实际测试后分析和评估不同算法的结果。他们将获得自动驾驶算法开发的知识和信心，通过出版物与他人分享他们的知识，并有可能选择自动驾驶汽车的研究和开发作为他们的职业道路。学生对智能移动领域和职业选择的看法将受到积极影响，从而增加对未来研究生学习的兴趣。测试的各种自动驾驶算法的评估和比较结果是对研究界的原创性贡献。这一结果将有利于自动驾驶汽车软件开发的进步，为社会带来减少污染、减少交通事故和拥堵、降低经济成本等效益。具体目标是：(1)为代表性不足的本科生提供经验，否则他们可能没有研究机会学习自动驾驶汽车开发的基础理论；(2)允许学生设计算法，在真实的测试课程中使用真实的车辆进行软件开发实践；(3)增强自信心、自我引导能力和研究技能；(4)增加对研究生课程感兴趣的学生人数，最终为工业界提供一支高素质的研发队伍。活动包括(1)每天密集的培训工作坊，用真实的车辆解决实际问题；(2)明确研究问题；(3)算法的设计、实现和测试；（四）对收集到的数据和结果进行分析和评价；（五）撰写技术报告和演示文稿；（六）项目前后的评估；（七）实地考察；(8)出版会后。本研究还通过识别在教学和学习自动驾驶算法中使用真实车辆的优点和缺点，以及使用真实车辆教授自动驾驶算法的最有效策略来提高本科教育，从而提高知识。该奖项反映了美国国家科学基金会的法定使命，并通过使用基金会的知识价值和更广泛的影响审查标准进行评估，被认为值得支持。

This project provides hands-on active learning opportunities for undergraduate students to conduct research for urban road self-driving functions using street-legal vehicles. It is uncommon for US undergraduate students to have opportunities to develop self-drive algorithms using real vehicles. In this research, students will analyze and evaluate the results of different algorithms after real-world testing. They will gain knowledge and confidence in self-driving algorithm development, share their knowledge with others through publications, and potentially choose autonomous vehicle research and development as their career path. Student perceptions about the smart mobility field and career options will be positively affected, leading to increased interest in future graduate studies. Evaluation and comparison results of various self-driving algorithms tested are original contributions to the research community. The results will benefit the advance of autonomous vehicle software development, bringing benefits to society including reduced pollution, less traffic accidents and congestion, and lower economic costs.The specific objectives are to: (1) provide experiences to underrepresented undergraduate students who otherwise might not have research opportunities to learn fundamental theories in autonomous vehicle development; (2) allow students to design algorithms to practice software development using real vehicles on real test courses; (3) strengthen their confidence, self-guided capabilities, and research skills; and, (4) increase the number of students interested in graduate programs and ultimately provide a quality research and development workforce to industry. Activities include (1) intensive daily training workshops with hands-on problem solving tasks with real vehicles; (2) defining research problems; (3) design, implementation, and testing of algorithms; (4) analysis and evaluation of the collected data and results; (5) writing technical reports and presentations; (6) assessments before and after the program; (7) field trips; and (8) post-meetings for publications. This research also advances knowledge by identifying advantages and disadvantages of using real vehicles in teaching and learning self-driving algorithms, plus the most effective strategies to teach self-driving algorithms using real vehicles in order to improve undergraduate education.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.