Monday 13 September 2010

Greenfoot and Junkbots

One of the criticism of the robot programming part of the Junkbots project is not everyone necessarily gets a go at the programming. To address this a new feature has been added to the project. There are now two parallel activities  as well as programming a robot; there is a separate programming exercise carried out at the same time which replicates some of the same actions of the robot but this time on screen.
Figure: Robot pushing a barrel

These exercises are based around the increasngily popular Greenfoot software (http://www.greenfoot.org/download/) which is free to download and use. This can be put on as many machines as are need enabling more people to have a go at programming.

The exercises initially gets participants to set-up the world, place a robot within it and get the robot to move across the screen. Building on the each previous exercise, the complexity increases and includes challenges (such as in the figure) where the robot pushes a piece of rubbish (in this case a barrel) off the screen.


Some of the material can be found at: http://www.computing.northampton.ac.uk/~scott/greenfoot_ex/sco1/default.htm. An example is shown below.







Tuesday 7 September 2010

A Virtual Environment Training System for Haptic Laparoscopic Surgery

Work from the collaboration between Dr Fengda Zhao  (Yanshan University, Qinhuangdao, P.R.China) and Dr Yinghui Zhang (Division of Computing, School of Science and Technology, University of Northampton, UK) is to be presented at  16th International Conference of Automation and Computing, University of Birmingham, Birmingham, UK, 11 September 2010.The work looks at developing a prototype haptics-based surgical training system


Most of the existing laparoscopic simulations which use simple objects are only for the purpose of the training of laparoscopic surgical skills. We have designed and developed an early prototype of a laparoscopic simulation system in which almost all of the soft tissue organs are modeled as deformable models which can be manipulated using a laparoscopic instrument with haptic feedback. Physics-based modeling is applied to realize collision detection, force rendering and elasticity deformation. A new mechanism to analyze and evaluate the injuries to the soft issues is introduced. Preliminary experimental results show that the simulator can meet the requirement of 1 kHz haptic loop update rate under complex virtual scene.


This collaboration was supported by a Leverhulme Fellowship .