Magic trick presented during scince week, with a robot assistant.

Student project by Simon Tomlinson. Program on board of a humanoid robot uses artificial intelligence technique (like A*) to determine next move in a game of incomplete information and team cooperation with adversaries (Spanish dominoes). It uses computer vision techniques to recognize tiles and movements to indicate the robot's decision. A GUI enables humans to participate remotely and a central server monitors the state of the game.

Poker playing robotic dog. Uses model-driven engineering with state machines and plausible logic.

Demonstrations of Task Planning and Motion Planning Integration.

Robot plans despite deterministic adversaries

This video shows our proposal for identifying objects by their shape and extracting the colors from within these shapes. The chosen shape recognition algorithm is the Histogram of Oriented Gradients. The complex objects we recognize are other humanoid robots (in particular Naos), similar in complexity for their anthropomorphic shape but less variable than human beings. This enables full learning of the environment colors in less than 1 minute on board of a Nao. See Vladimir Estivill-Castro and Jordi Radev: Humanoids Learning who are teammates and who are opponents The 8th Workshop on Humanoid Soccer Robots at 13th IEEE-RAS International Conference on Humanoid Robots

This video displays the progress of the MiPal team from Universitat Pompeu Fabra in Barcelona Spain and Griffith Univeristy in Brisbane Australia. This video was presented for qualificaiton in RoboCup 2015

This video displays the research on integrating multi-platform planing in the MiPal infrastructure by Jonathan Ferrer at Universitat Pompeu Fabra (Barcelona, Spain) and the logic-based finite-state machines tools developed by Robert Coleman and Carl Lusty at Griffith Univeristy in Brisbane Australia. It also has highlights of MiPal's participation in RoboCup 2012. .

This video displays the MiPal facilities in Brisbane Australia, and Nao robots behavior built using Model Driven Engineering. The behavior is synthesized from finite state machines. Sonar positioning without visions as well as team self-positioning on a filed where both goals are yellow. For computer vision, we use opencv for line recognition and we distinguish line shapes (associated document can be found here).

This video displays the MiPal facilities in Brisbane Australia, and Nao robots playing soccer after Model Driven Engineering. The behavior is synthesized from a formal logic and finite state machines. A game controller is also ported to different platform including and iPad

MiPal goals in RoboCup 2006 held at Bremen Germany

MiPal demonstration of symmetry image recognition to distinguish the event of a kick in RoboCup 2005 held at Osaka Japan

MiPal goal in RoboCup 2004 held at Lisbon Portugal

MiPal goals in RoboCup 2003 held at Padova, ITALY

German Open 2003 (Griffith-Blue vs Bremen Read : Final score 1-2)

This is an Alderbaran Nao and a Lego NXT running the same model to illustrate Model-Driven Engineering, same model, runnign in two paltfrom. The model is a series of finite-state machines with transitions labeled by a common sense logic. The models are also verified using formal model checking.

This is an Aldebaran Nao simulating a Microwave. Using requirements engineering, the state machine that runs the microwave has been compiled automatically from the high level design elicited from the natural language description of the requirements.

This video shows a NXT configured as a Mine Pump controlled by a model of finite-state machines with transitions labeled by expressions of a common sense logic. The model is directly interpreted for validation, the model is also verified using model-checking techniques. All the machines are organised as a vector and a round-robin schedule is used to run them concurrently, but the result is a sequential program.