IIT Madras: Researchers Develop Motion Planning Algorithms Capable Of Thinking Like Human


Sonam Rana updated logo

Sonam Rana updated Content Curator updated

Content Curator updated

The main advantage of this algorithm is that it will significantly improve efficiency of computation over various other well known motion planning algorithms.It will lead to strong applicability of GSE-based algorithms during time-sensitive planning scenarios.

The Researchers of the Indian Institute of Technology (IIT),Madras have developed a class of more efficient and faster ‘motion planning’ algorithms that can think just like human beings.These algorithms will also enable autonomous aerial,surface(ground) vehicles to navigate in obstacle-cluttered environments. 

IIT Madras : Admissions

The team claims that the algorithms have been developed on a novel motion of GSE(Generalised Shape Expansion) that enables planning for a safer & dynamically feasible trajectory for autonomous vehicles.All these approaches have been found to yield superior results compared to various existing seminal & state-of-the-art motion planning algorithms.

Due to its novel 'safe' region calculation, it provides a crucial advantage of planning during time-sensitive scenarios arising in applications like Self-driving cars, ISR operations,Planetary exploration, Disaster response,Aerial drone delivery and various others, the research team said.

This Research was led by Satadal Ghosh, Assistant Professor, Department of Aerospace Engineering, IIT Madras who has published several research papers in internationally reputed journals such as AIAA Journal of Guidance, IEEE Control Systems Letters, and top-tier conferences like AIAA SciTech, IEEE Conference on Decision and Control (CDC) etc.

The team also included ; Vrushabh Zinage (IIT Madras Alumni) who is currently a doctoral research scholar at University of Texas Austin (USA), Nikhil P, an Analyst in Goldman Sachs & Adhvaith Ramkumar who is currently a Graduate student at Warsaw University of Technology, Poland

These GSE-based algorithms function by calculating a 'safe' region that consists of large 'visible' areas in the environment, customised to ensure navigability. Following this, they select a random point in this 'visible' region & connect it through a safe 'edge' to safely reachable regions discovered so far.Eventually, these algorithms can almost always connect any two points in any environment that satisfy certain basic criteria," Zinage said.

IIT Madras : Distance Education

Briefing about the Current Status of this research,the team said that it is currently limited to theoretical development and improvement of the GSE-based algorithms along with extensive realistic simulation-based validation of the same. The team is further planning to perform actual implementation of these algorithms on unmanned aerial & Surface vehicles in the near future.

"In case of dynamic environments, where environmental knowledge is just limited to information from on-board sensors or when mission commands intervene in the vehicle movement because of dynamically evolving mission-critical needs,for example Intelligence,Surveillance and Reconnaissance operations or planetary exploration using rovers, frequent time-critical re-planning of motion is usually called for.

"Even in such cases,our present study suggests that on-the-go motion planning becomes significantly easier by our GSE-based algorithms because of the unique nature of visible regions calculated by these algorithms at different points in the environment," the Research leader, Satadal Ghosh, said.

Also Check : 

IIT Madras, Tata Power sign MoU to Collaborate on R&D, Training, Tech Solutions: Read More Details Here

Category:

Comments


No Comments To Show