Educational Robotics in a Systems Design Masters Program

ICALT‘2003 AthensUwe Gerecke Educational Robotics in a Systems Design Masters Program

Educational Robotics in a Systems Design Masters Program

Uwe Gerecke, Patrick Hohmann, Bernardo WagnerLearning Lab Lower Saxony (L3S), Hanover, Germany

{gerecke,hohmann,wagner}@learninglab.de


Talk Outline

• Educational Roboticsn Robot Platforms in Education

• Modular Educational Robotic Toolbox n What is MoRob about? n Project Aimn Topics

• Robot Curriculum + Project-based Robotics Courses n Legon MSc. Systems Design at University of Hannover

• Evaluation n Testbeds, Progress, First Insights

• Examples• Conclusions


MoRob – Modular Educational Robotic Toolbox

A project to develop a framework for teaching/research with Robotics

Project Partners:• L3S Hannover• KTH Stockholm, Centre for Autonomous Systems• Stanford University, Robotics Laboratory

Funding: WGLN – Wallenberg Global Learning Network


Autonomous Mobile Robots

In recent years, autonomous mobile robots have..• become a very popular topic • gained high visibility in media (TV, newspapers, internet)• been used increasingly in research, entertainment and

industry

Going along with this..• increasing number of university robotics courses• increasing number of dissertation projects


Educational Robotics (I)

Robotics offers an excellent basis for teaching a number of different engineering disciplines and their integration into systems.

For example, courses with robots on:

• Software engineering projects [Gustafson, 1998]• Data structures courses [Dannelly, 2000] • Genetic programming [Bianchi and La Neve, 2002]• Hardware based vision [Bianchi and La Neve, 2002]• Artificial Intelligence [Paul, Hafner, and Bongard, 2000]


Educational Robotics (II)

Robots can be used as a tool in many different disciplines

• Teaching/Demonstrating

• Exercises/Labs

• Research projects (PhD. level)


Robots in Teaching

In student projects, exercises help to develop engineering skills:• Creativity• Teamwork• Designing• Problem solving

Teaching the technical content could very easily be done in student teams through problem-based learning.

This improves and increases the attractiveness of learning.


Reasons to Use Robots

• Robots as demonstration tools: abstract concepts can be turned into real-world problems and solutions [Kumar and Meeden, 1998]

• Motivation improvement when real-world objects are included [Garcia and Patterson-McNeill, 2002]

• Motivational factor to attract high-school students [Oppliger, 2001]

• Robots are very popular and stimulating for university students [Wang, 2001]


What is MoRob about?

Universities need robot platforms which• are flexible and modular, yet powerful• can easily be customized to the requirements of different

subjects and research goals

MoRob project aims:• develop such a platform• provide standard set of robot control modules + teaching units

This is aimed at teaching undergraduate and graduate students as well as for PhD research.


Why do we need MoRob?

Common problems in educational robotics:• Use “primitive” systemà limited platform

• Construction of own systemà effort requires many resources

(implementation, maintenance)

We do not want teachers and students to re-invent the wheel all over again!


What about Lego?

LEGO Mindstorms: • Provides a flexible and easy to use platform• Reaches its limitations very easily

(physically, programming power, sensors,scalability, etc.)

So: Lego is not the best platform to be used in advancedcourses.


Educational Robotics Requirements (I)

• Flexible architecture which considers mechanical, electrical and software interfaces

• Layered architecture that allows interfacing to the system at all levels from motors to advanced behaviours (comprehensive API required)

• Comprise a variety of basic components to allow for construction of basic mobility systems (e.g. sensor modules)

• Library of standard modules for navigation, detection of obstacles, etc.

• Simple set of interfaces to be used with MATLAB and Java

• Comprehensive suite of documentation


Educational Robotics Requirements (II)

The platform should be• easily configurable for a particular course• interfaced to standard educational tools such as MATLAB, Java and

C/C++• sharable across a number of institutions

Characteristics summary:• Scalable performance• Modular setup • Flexible interfaces• Easy configuration


MoRob Overview

• Development of a Scalable Processing Box (SPB) • Definition of a modular robot model curriculum

+ Applications in Problem-Based Learning• Development of associated course materials• Demonstrator modules (Toolboxes)• Provision of an Internet framework (Repository)• Evaluation of results


Scalable Processing Box (SPB)

Core of our system: SPB, a standardized processing unit for easy experimentation with any kind of robotic platform.

• Standardized• Scalable• Modular• Several SPBs can be combined


Educational Robotics Processing Requirements

• Simple handling of the SPB• Easy configuration• Robustness from hard- and software's point of view • Real-time capability• Understandable and praxis-oriented documentation and

examples

The range of users to work with the SPB will reach from studentsdoing their first steps in the subject of robotics to people using the SPB as base system for their scientific work.


Curriculum Development

Essential part of MoRob: provide a model curriculum for using robots in teaching and exercises.

• Choice of teaching units and exercises in different subjects• Choice of approaches at different levels• In particular: promotion of Problem-Based Learningn increases the attractiveness of learning while performing student

projects through learning by doingn supports the development of engineering skills like creativity,

teamwork, designing and problem solving


Associated Course Materials

Development of associated course materialsn lecture notesn slidesn videosn exercisesn solutions

• To supplement the platform• Provision of a variety of course materials • Cater for a number of different courses within the Robotics,

Computer Science and Engineering curriculum• To be easily adapted to a variety of courses


Development of Toolboxes (Demonstrator Modules)

Teaching of robotics classes requires a platform to demonstrate key concepts (different modules for different concepts).

We need: library of standard modules for navigation, detection of obstacles, basic trajectory following, etc.

Sample module: demonstrator for localization experiments.

This demonstrator allows:n working with students in the labn evaluation of pedagogical meritn testing of software architecture and complete system


Internet Framework

• Internet repository for sharing of materialn course materialsn toolboxesn curriculan softwaren etc.

• WWW-portal for course material and technical documentation• Provision of a tutorial help system• Remote control of the robot through wireless LAN


Robot Platforms in Education

The quality and properties of a robot platform determine which concepts can be conveyed to the students.

Platforms can be classified into three categories:1. Introductory level, schools (eg. Lego Mindstorms, Rug Warrior)2. Undergraduate level (eg. Evolution ER1)3. Advanced robots for research (eg. Pioneer 2AT, RobuCar)


Platforms at L3S


Lego Mindstorms


Evolution ER1

•New platform launched in 2002• Constructed from a modular set

of aluminum profiles • Laptop• SDK under MS-Windows• Limited with respect to sensing

facilities and software flexibility• Extensible


ER1 Sensors

• Camera• Optional 2nd camera• Wheel encoders• Optional Infrared

• Other groups: e.g. ultrasoundsensors


Pioneer 2AT (ActivMedia)


Pioneer 2AT Hardware

• 2AT base unit

• Embedded system running

Linux/RTAI

• Sensors:- Laser scanner- GPS- Compass- Gyroscope- Wheel encoders


Robot Curriculum + Project-Based Robotics Courses

Mini-Project “Autonomous Service Robots” (successfully completed)

n project-based learning with Legon videon evaluation

New lab project in summer semestern localization via access pointsn Institute of Communications

Engineeringn project-based taskn ER1

MSc. Systems Designn starting autumn 2003n 3 projects


Mini-Project “Autonomous Service Robots”

• Hardware laboratory 5th semester BSc. “Applied Computer Science”• Set exercises, then Mini-Project over 4 weeks• One choice: “Autonomous Service Robots”• Application scenario: burning oil springs (build fire-fighting robot)• First project to be filled• Teams: equipped with Lego Mindstorms

kit and LeJos• Self-organized tasks, little interference

by tutor• All groups achieved goals successfully• Initial evaluation


Lab Project “Localization with Bluetooth”

• Localization task• Use signal strength of Bluetooth

access points• Institute of Communications

Engineering• Project-based task• ER1


MSc. Program “Systems Design”

• Start in autumn 2003 at Hanover University• Targeted at CS graduates with strong software experience• “System“: mobile phone, engine controller, mobile service

robot, etc.• Series of robotics projects (3 consecutive semesters) as an

essential part of the degree• Teams of 4 individual students

• Testbed for SPB• Testbed for evaluation


MSc. Structure and Practical Work

Block I: Fundamentals (12 CP) Block IV: Soft Skills (14 CP)Block II: System Design (28 CP) Master Thesis (30 CP)Block III: Practical Work (36 CP)

6 CPs Labs/Seminars 6 CPs Design Skills

9 CPs Synthesis 2 (Implementation) Project III

9 CPs Synthesis 1 (Design) Project II 6 CPs Analysis/Modelling Project I


Practical Work / Projects

Project I: Systems analysis and modelling of a systemn Matlab/Simulink to create a simulation model of an existing systemn software, electrical and mechanical components, static and dynamic properties

Project II: Synthesis and designn Prototype as a virtual modeln Same tools from project I (Matlab)

Project III: Implementation and integrationn No Matlab, implemented in the real worldn Software developmentn Electrical devices (sensors)n Mechanical construction


Evaluation (I)

• We intend to provide informationas to the suitability of projectlearning for engineering education concerning robotics and using our MoRob system.

• To answer these questions: comparative evaluation studies.


Evaluation (II)

Our knowledge interests:

• Does MoRob and the project method improve student learningin robotics, mechatronics and autonomous systems?

• Is the system useful to integrate research and scientific methods better and faster in student teaching?

• What is the best way to supplement the project and problem based learning by courses?

• Can MoRob initiate collaboration between students from different countries?


Evaluation (III)

Evaluation of the learning processof the students:n Motivationn Creativityn Knowledge gainsn Collaborationn Theoretical concepts usedn Precisionn Quality of resultsn How do the students cope with technical challenges?n Understanding and use of scientific and engineering methods?


Evaluation Testbeds

Main part of evaluation:• New MSc. Systems Design at Hanover• New international MSc. Robotics at KTH• Robotics Class 225 at Stanford

First year evaluation:• Mini-Project "Mobile Service Robotics" in Hanover, Feb. 2003• Autonomous Systems course at KTH, June 2003• Experimental Robotics class at Stanford, June 2003


First Insights (I)

• Within the problem-based learningtask, students could bring in theirown ideas to a large extent and hadto acquire new knowledge on their own

• The tutor only had to give little directions and students could organize their work freely

• Teamwork and communication skills were felt to be among the most important things learned during this project

• Many students felt that this was the best or most interesting project work they had ever done


First Insights (II)

• Overall, the students judged theLego platform and its environmentas positive and appropriate

• However, there were a large number of individual comments highlighting the shortcomings of this platform, for example: imprecise sensors, limited software, small memory, problems with multithreading and arithmetic

• Students were highly motivated on this course and usually achieved more than was expected


Conclusions

• MoRob: robot platform for university teaching and research• Scalability, modularity, flexibility and ease of use• Exercises with robots help to develop engineering skills like

creativity, teamwork, designing and problem solving• MSc. in Systems Design: problem-based learning in robotics• Students will have the tools to tackle complex design tasks• Broad range of engineering concepts can be taught• Vehicle for experimentation in a variety of courses• Standardized teaching environment for different applications


Questions?




Pioneer 2AT Inside


Pioneer 2AT Sensors


The L3S from a Robot’s Point Of View


Videos

Lego Mini-Project Exploration with 2AT

Documents

Educational Robotics in a Systems Design Masters Program