This master class prepares for a new type of engineering discipline: design and creation of intelligent systems, products, and related services. The class will be a unique opportunity to get familiar with a number of powerful conceptual and intellectual tools to understand and create adaptive behavior at a system level. The context of Lego is chosen because it is already an example of a system. But in this class the creative goal is to make a leap forward, extending the scope of the existing system such that adaptive behavior becomes the central theme. The new Lego should be equally fascinating for grown-ups and children, women and men.
Target Competency Area(s), Learning Objectives and Level(s)
System level thinking has had and still has an enormous impact upon the development of technology. When working at a system level one does not study individual component behavior, such as Ohm’s law for an electrical component; instead one addresses bigger questions such as the stability of the feedback loops, information throughput, or learning capacity. The learning objective includes classical control, reinforcement learning and adaptive control, pattern recognition and principles of Lego. The teachers will offer knowledge, theory and information. The students are invited to study this and demonstrate their understanding of the information through something that they create. The teachers will encourage depth, through theory offerings, tools and backgrounds provided. They will insist on solid arguments: in materials, in sensors, in form creation, and in control. The students, working in groups, create at least one new sensor and one new actuator, for example a body-computer-interface sensor and a controllable pneumatic linear actuator. Using this, next to the existing Mindstorm components, a fascinating demonstrator is designed, perhaps a piece of electronic art or a life-like creature.
Lectures and workshops:
- classical control (Delbressine, Feijs)
- reinforcement learning and adaptive control (Ahn, Barakova)
- pattern recognition (Hu)
- principles of Lego (Bartneck, Delbressine)
- guest lectures
Component level design activities:
- innovative sensor
- innovative actuator
System level design activities
Fascinating demonstration system including applications of the theory and the newly designed components.
The main deliverable of this project is an extension pack to the current Mindstorms NXT set. This extension set forms a system of at least one new sensor, actuator and Lego brick. The Lego brick may be manufactured using our 3D printer and should incorporate a new mechanical principle.In more detail, the deliverables are:
• working model of the extension pack
• report in MS Word describing the extension pack
• technical documentation of the extension pack, such as technical drawings and UML models
• photos of the extension pack
• movie of the extension pack in action uploaded in high quality to SurfMedia and available for download from there.
List of Available Reference and or Background Materials
- Tutorial on how to install and run Java on Lego Mindstorms NXT using Eclipse
- Lego hardware and software developer kits
- Lego Java
- Lego LabView
- Brick Journal
- Building Robots with LEGO Mindstorms NXT by Mario Ferrari, Guilio Ferrari, and David Astolf
- Extreme NXT: Extending the LEGO MINDSTORMS NXT to the Next Level (Technology in Action) by Michael Gasperi, Isabelle L. Hurbain, and Philippe E. Hurbai
- The Unofficial LEGO Builder’s Guide by Allan Bedford
- Maximum Lego NXT: Building Robots with Java Brains by Brian Bagnall
- Building Robots with LEGO Mindstorms NXT by Mario Ferrari, Guilio Ferrari, and David Astolfo
- The LEGO Mindstorms NXT Idea Book: Design, Invent, and Build by authors of thenxtstep
- HiTechnic Products
- Lego Community
Dr. Christoph Bartneck – I am an assistant professor in the Department of Industrial Design at the Eindhoven University of Technology. I have a background in Industrial-Design and Human-Computer Interaction and my projects and studies have been published in various journals, newspapers and conferences. My interest are in the area of Social Robotics, Design Science and Multimedia Applications. I previously worked for several companies including the Technology Center of Hannover (Germany), LEGO (Denmark), Eagle River Interactive (USA), Philips Research (Netherlands) and ATR (Japan).
Prof.Dr.ir Loe Feijs – I am a professor holding the chair Industrial Design of Embedded Systems. I studied electrical engineering and have a Ph.D in computer science. I worked in telecommunication research (CSELT), telecommunication industry (PTI), in software research (Philips Nat Lab), in embedded systems (EESI), and in industrial design (TU/e). I was the director of the Eindhoven embedded systems institute. With co-authors I wrote several books on formal specifications and created mathematical theories of modular software design, such as lp calculus and relation partition algebra (RPA). Moreover I enjoy hands-on work, witnessed by my 4D- sketching, Mondrian software, the assignment Minimax and the results of DA201.
Dr. Emilia Barakova – My research interest lay in information integration at different levels of organization: (multi) sensory, perceptual, contextual and behavioral is integrated for complex (intelligent)task solving. The complex or intelligent task is considered from the behavioral perspective. The behavioral setup facilitates the investigation from neurobiological as well as from autonomous agent points of view. At present I am studying how contextual, behavioral and sensory information find a common representation in a model for spatial navigation, inspired by experimental findings of hippocampal functioning. Novelty and familiarity gating plays an important part of this study. Tests in a robotics setup are intended to illustrate the functionality of the behavioral model. Having as a testbed the navigation setup, I am trying to find a balance between neurobiological inspirations and computational soundness in my modeling and experimental work.
Frank Delbressine, Hu Jun, Rene Ahn