CS666: Mobile Robotics (M.Tech. /Ph.D. only Elective)
Instructor: Prof. Shivashankar B Nair
The course serves an exercise to know methodologies to bridge the ‘soft‘ world with its ‘physical‘ counterpart. The major focus of the course is to embed intelligence into connected physical entities such as embedded controllers and mobile robots. Since intelligence needs to generated by each entity in this dynamic network, algorithms used to churn out intelligence need to be mostly decentralized. This calls for populations of learning algorithms – either homogeneous or heterogeneous - that interact with one another within this network and share their respective information and intelligence. Such mechanisms are often seen in Nature. The algorithms discussed in this course are thus, those inspired from Nature. The course also includes the description of robots and their constituents, such as sensors, actuators and controllers.
The course is lab. intensive and deals with the use of real and connected robots such as the Legos and Firebirds. It also exposes the use of the Raspberry Pi as one of the controllers. Decentralization of intelligence is achieved using an open source Multi (Static and Mobile) Agent platform nicknamed Tartarus, developed at the Robotics Lab. of this Dept.
For conducting robot(s) based experiments, we shall be using a state-of-art robot simulator called, Webots. Additionaly, we will also endeavour to expose students to real robots and related hardware via the Internet using interfaces recently developed at the lab. using Tartarus.
- Theory sessions: Includes an introduction to mobile robots and related hardware; Nature-inspired and population mechanisms and algorithms such as Evolutionary algorithms, Artificial Immune Systems, Ant Colony based algorithms, Particle Swarm optimization, etc. and their use in the realm of mobile robotics.
- Lab sessions
- Introduction to Tartarus: A brief introduction and hands-on on Tartarus Multi-Agent System and its underlying language.
- Introduction to Webots
A brief introduction to the Webots robot simulator together with hands-on sessions
The course involves a rich set of assignments aimed at testing and improving the skills learnt across the course.
The course will conclude with projects involving mobile robots employed for a real world task. Some of the past projects under the course include multi-robot maze solver, decentralised dancing robots, telemetry control of real robot, etc.
In addition to the Webots simulator, the inclusion of actual hardware into the course is also in the pipeline. Since the current semester is to be conducted online, the same will be facilitated over the Internet, thereby allowing students to control robots and other hardware, remotely.
Since hardware available at the lab. is limited, we generally limit the number of registrants to 20 (M.Tech. & Ph.D.).
Register for this course if you have a passion to learn and program for real and open world problems.
For more information please feel free to connect, in addition, to the Research Scholars Ms. Divya Kulkarni (email@example.com), Mr. Suraj Pandey (firstname.lastname@example.org) and Ms. Menaxi Bagchi (email@example.com), at the lab.