EV Mechanica

Subscribe to EV Mechanica's Current Newsletter & never miss an update!

    Close Menu
      Facebook X (Twitter) Instagram
      EVMechanicaEVMechanica
      EVMechanicaEVMechanica
      • Home
      • News
        • E-Mobility
        • EV Battery
      • Charging Stations
      • Policy
      • Research
      • Interview
      • Jobs
      • Events
      • E-Mag
      • Subscription
      Facebook YouTube LinkedIn WhatsApp
      EVMechanicaEVMechanica
      Home » IIIT Hyderabad’s AI-Driven Car Revolutionizes Robotic Navigation

      IIIT Hyderabad’s AI-Driven Car Revolutionizes Robotic Navigation

      Rashmi VermaBy Rashmi VermaNovember 19, 2024Updated:November 19, 2024 E-Mobility 3 Mins Read
      IIIT Hyderabad's AI-Driven Car Revolutionizes Robotic Navigation
      Share
      Facebook Twitter LinkedIn WhatsApp

      IIIT Hyderabad’s Self Driving Car is an electric vehicle that performs point to point Autonomous Driving with collision avoidance capabilities over a wide area. Equipped with 3D LIDAR, Depth Cameras, GPS systems and AHRS (Attitude and Heading Reference System which essentially means sensors on three axes to estimate its orientation in space), the car can also accept Open Set Natural Language commands and follow those commands to reach a desired destination. SLAM-based point cloud mapping is used to map the campus environment and a LIDAR-guided real-time state estimation allows for localization while driving.

      Open Set Navigation

      To better understand open set navigation commands, let’s first consider how humans often navigate – with minimal map usage, relying on contextual cues and verbal instructions. Many navigation directions are exchanged based on recognizing specific environmental landmarks or features, for example, “Take right next to the white building” or “Drop off near the entrance”. In a parallel context, autonomous driving agents need precise knowledge of their pose (or localization) within the environment for effective navigation. These are typically achieved using high-resolution GPS or pre-built High-Definition (HD) Maps like Point Cloud, which are compute and memory-intensive.

      IIITH’s Efforts

      They address it by exploiting foundational models that have a generic semantic understanding of the world which can be distilled for downstream localization and navigation tasks. This has been achieved by augmenting open-source topological maps (like OSM) with language landmarks, such as “a bench”, “an underpass”, “football field”, etc. which resemble the cognitive localization process employed by humans. These enable an “open-vocabulary” nature that allows navigation to places for which the model is not explicitly trained, leading to a zero-shot generalisation to new scenes.

      Differential Planning

      Mapping, localization, and planning form the key components in the Autonomous Navigation Stack. While both the modular pipeline and end-to-end architectures have been the traditional driving paradigms, the integration of language modality is slowly becoming a defacto approach to enhance the explainability of autonomous driving systems. A natural extension of these systems in the vision-language context is their ability to follow navigation instructions given in natural language – for example, “Take a right turn and stop near the food stall.” The primary objective is to ensure reliable collision-free planning.

      USP: NLP+VLM
      To achieve this capability, IIITH has developed a lightweight vision-language model that combines visual scene understanding with natural language processing. The model processes the vehicle’s perspective view alongside encoded language commands to predict goal locations in one-shot. However, these predictions can sometimes conflict with real-world constraints. For example, when instructed to “park behind the red car,” the system might suggest a location in a non-road area or overlapping with the red car itself. To overcome this challenge, they augment the perception module with a custom planner within a neural network framework. This requires the planner to be differentiable, enabling gradient flow throughout the entire architecture during training which eventually improves both prediction accuracy and the planning quality. This end-to-end training approach with a differentiable planner serves as the key sauce of our work.

      AI-Driven Car Autonomous Navigation Stack gps High-Definition IIIT Hyderabad Point Cloud Robotic Navigation
      Share. Facebook Twitter LinkedIn WhatsApp
      Rashmi Verma

      More article from Rashmi Verma

      Keep Reading

      Bihar Set to Deploy 150 Electric Buses August

      Tesla Unveils Oasis Supercharger with Solar, Off-Grid Batteries

      Heybike Galaxy series debuts with sleek, powerful e-bikes

      Leave A Reply Cancel Reply

      two × three =

      E-MOBILITY

      Bihar Set to Deploy 150 Electric Buses August

      July 4, 2025

      Heybike Galaxy series debuts with sleek, powerful e-bikes

      July 4, 2025

      Mahindra launches XEV 9e Pack Two in India

      July 4, 2025

      Matter aims to sell 10,000 electric motorcycles in 2025

      July 4, 2025

      Articles

      Delhi’s Fuel Ban Flip-Flop: A Policy Failure in Motion

      Introduction: A Capital Confusion Delhi government took a U-turn on its decision to implement the…

      India’s Largest Electric Bus Procurement Signals a Green Public Transit Revolution

      The Government of India has introduced the largest electric mobility project to date the procurement…

      Delhi’s Vehicle Age Ban: A Harsh Road to Clean Air or Policy Misfire?

      The government took fairly bold action to address Delhi’s chronic air pollution crisis by implementing…

      © 2025 EVMechanica.com.
      • Home
      • About Us
      • Contact Us
      • Subscription

      Type above and press Enter to search. Press Esc to cancel.