The iRobot Create // Entry 5
During my analysis I honed is the Roomba Create series as the robot that I want to focus on. As mentioned in the report, this is the variant that iRobot created in order to make the process of hacking and working with the Roomba as accessible and simple as possible while also reducing the entry cost so that anyone could buy and do things with it. Its movement ai, learning and sensors make this an extremely valuable research tool, especially when it comes to navigation and movement robotics.
Some references that cover educational application:
Dekan, Martin, et al. "iRobot Create Used in Education." Journal of Mechanics Engineering and Automation 3.4 (2013): 197-202.
Tribelhorn, Ben, and Zachary Dodds. "Evaluating the Roomba: A low-cost, ubiquitous platform for robotics research and education." Proceedings 2007 IEEE International Conference on Robotics and Automation. IEEE, 2007.
Ekong, Donald U. "Resources for using robotics to inspire interest in engineering." IEEE Southeastcon 2009. IEEE, 2009.
Other existing platforms:
Mondada, Francesco, et al. "The e-puck, a robot designed for education in engineering." Proceedings of the 9th conference on autonomous robot systems and competitions. Vol. 1. No. LIS-CONF-2009-004. IPCB: Instituto Politécnico de Castelo Branco, 2009.
Huq, Rajibul, et al. "QBOT: An educational mobile robot controlled in MATLAB Simulink environment." CCECE. 2009.
The evaluation component of this assessment is something that I’m not feeling overly familiar with (definitely something that I need to talk with Ricardo more about). This is mostly due to the fact that I had another commitment half way through the lecture, but from what I can tell this comes down to figuring out a way to test an aspect of my proposed bot. No surveys as that can lead to very inconsistent and useless results, but something more tangible.
I want to tackle this from two perspectives, the robot and the software. Designing an interface or contact point for the users to interact with the robot on a deeper level as well as looking a programmable robot. I would also like to focus on this from an educational standpoint, academic and also at a novice level such as high school classrooms.
Software needs to be adaptable and applicable at both levels of education and research. Possibly two different coding methods, MIT scratch integration as well as hard code for those that understand how to use it. Easy to navigate, easy to connect with the bot, easy to update bot software, easy to revert bot software, possible safety barriers to prevent irreversible damage.
The robot needs to be relatively powerful, but cost effective. So not the most high tech stuff involved, for example the latest 3D cameras or something along those lines, but perhaps support for peripherals such as the Kinect (definitely something to research and look at how people are getting around this issue). I’m also thinking something relatively small in comparison to the Roomba, mainly due to space saving and cost (micro drones compared to super high end drones). Possible interface on the robot itself to allow for minor, on the go changes or switches in behaviors. Obviously room to expand and integrate extra computational power / sensors with devices such as the Arduino or components like sonar sensors.
I think a lot of this will be more solid when I take a slightly deeper look into how iRobot has changed the Create over its iterations based on user demands and also how the competition has responded to its presence in the market.
1) Wireframing/hifi prototype of software interface and taking this to users to see if it is easily usable and understandable.
2) Paper prototypes. Used at work.
3) I actually quite liked the idea of acting out a robot. Could possible do an experiment relating to how people used a software and then the results it would give when executed.
4) Actual models of the components of the robot? Things like a top panel with the interaction elements or the robot as a whole to get an idea of the size.
Will definitely need to have a look at what people are doing and have a chat to Ricardo about this to get a better idea of how I want to execute this.