Robotworld
|
Principal Investigators:
Prof. John Williams - Dept. of Civil and Environmental Engineering
Additional Investigators:
Prof. Alex Slocum - Dept. of Mechanical Engineering
Prof. Marty Culpepper - Dept. of Mechanical Engineering
Project Overview
Robot World will systematize the deployment and running of courses
that use design projects as a vehicle for teaching engineering.
Project Based Learning courses pose teaching and logistic challenges
in that they typically use hardware and software laboratories and have
student teams that operate with a high degree of autonomy. The initial
focus of Robot World will be on four MIT courses that use the building
of robots and robot software to motivate students to learn various
aspects of engineering. This project will take an engineering product
development approach to assembling customer needs and deploying a
range of Web Services to meet those needs. We believe that the lessons
learned during creation of Robot World will scale to the broader
engineering community envisaged in Engineering World.
Good design does not just happen, it is planned and executed in a
systematic way, with random "Aha!" thoughts interspersed as
catalysts to the creation of the product. While students need a high
degree of autonomy they nevertheless need both technical and time
management guidance in order to pass through the necessary milestones
on schedule. Ideally, courses also need to provide
"apprenticeship" experience for the student so that they can
"see" how good designers function.
Robot World will be initially deployed and tested in two mechanical engineering courses where
students are given a broad ill-defined task, such as "Design a
robot to compete in a competition," which encourages them to
explore different strategies and concepts using analytical, simulation
and hands-on experiments within a bounded design space. While these
courses have a single overarching goal, they are typically organized
into a series of knowledge check points that must be attained in order
to reach the final goal. Each check point represents an integrated,
but limited, piece of knowledge that must be grasped and understood
and then utilized to make a design decision. A key need in project
courses is to keep the student on track and take action when the
student falls behind the planned trajectory as illustrated below.
To optimize the student’s learning and enable the advisors (human
or online) to help, Robot World will systematize the following
steps:
- Present the student with timely, personalized instruction (both lectures, online content and off-campus experts)
- Assess the student’s progress, and raise "red flag" warnings as warranted
- Advise the student about the steps necessary to get back on track, including marshalling the necessary human and online resources to facilitate recovery, including possibly showing the student how a good designer would approach the problem.
The learning process we propose requires that advice and content be
customized to the student’s needs. This implies that at any given
stage we must (1) identify the student’s needs e.g. categorize the
student and assess their understanding of the subject material, and
(2) select and marshal appropriate material, customized to those
needs, from a content store and similarly marshal and select the human
resources.
We propose to make two innovations that will improve the learning process.
- Generalize the concept of the Design Notebook. Traditionally
assessment in design courses is done not only by tests and quizzes but
by also requiring the student keep a "Design Notebook." By
moving the student’s Design Notebook, test results, and other
accumulated work online, we can make the human advisor’s job easier
and also lay the foundation for automated Advisors in the next phase
of the project. We also note that if the student’s portfolio is
online, we can bring outside design experts (MIT alumni or industry
partners, for example) "into" the classroom.
- A significant effort in this project will be to package the
material already available in courses 2.007 and 2.000 into a
form that can be customized and filtered for the individual student.
The material will be categorized so it can be presented with different
levels of detail: the complete neophyte may choose to follow the
design process plan in fine detail, while the more experienced student
can stick to the outline on a higher level. For example, on the
highest level the design process says that once the product type/area
is defined, 1/3rd time is allocated for problem identification and
strategy and concept generation, 1/3rd time for detailed engineering,
and 1/3rd time for build and test. Thus, Robot World helps students
learn that solving a technical problem requires managing technology,
time, and resources, and the ability to learn new things and
accumulate information in a useful manner.
The Robot World system will be tested in a two MIT courses, 2.000 and 2.007,
and these will provide initial feedback on the prototype system.
The results of these tests will be developed
as case studies that identify best practices, including a set of
templates for replication in other courses at MIT and at other
schools. Plans for scale-up in year 2 include deployment in Cambridge
University in the UK and in year 3 in Singapore and Japan.
For More Information: