Hardware Podcasts

Episode 85: Drone vision and navigation - Margarita Chli

Robot Talk · 17 May 2024 · 25m · Episode 85

What this episode covers

Margarita Chli, Professor of Robotic Vision at the University of Cyprus and ETH Zurich, discusses her lab's research into vision-based navigation for small aerial drones. She explains how her team uses a mix of machine learning and handcrafted geometric methods to handle perceptually degraded environments like smoke and rain. Chli details the architecture of their collaborative drone swarm system, which uses centralized and peer-to-peer communication to map environments and correct localization errors. She highlights applications in search and rescue, wildfire monitoring, and infrastructure inspection, emphasizing the need for robust perception in hazardous conditions.

Most relevant to

early-stage companies

Topics

Mentioned

Companies
University of Cyprus, ETH Zurich, Google
Also
Fukushima, Parthenon

Related episodes from other shows

Other shows on the same ground, closest first.

  • The Robot Report Podcast · 1h 10m

    State of robotic perception with RGo Robotics' Amir Bousani

    Amir Bousani, co-founder and CEO of Argo Robotics, discusses the company's exit from stealth mode with $20 million in funding and its perception engine for mobile robots. He explains how Argo's vision-based system enables 'human-level perception' to identify objects and navigate complex, unstructured environments where traditional LiDAR fails. The conversation covers the limitations of current robotic sensing, the strategic choice of camera-based hardware reference designs, and the future of sensor fusion in robotics.

    Sensors & measurement in productsAI & machine learning
    7 chapters

  • Embedded.fm · 1h 06m

    301: Giant Novelty Check

    Carter Frost, president of the Cabrillo College Robotics Club, explains how his team won the 2019 NASA Swarmathon by solving critical localization issues that defeated other universities. The team overcame hardware flaws in the Swarmy robot, specifically IMU magnetometer interference from motors and GPS antennas, by developing a custom soft/hard iron calibration routine. They also implemented a correlated random walk algorithm and drove slower to avoid obstacles, allowing them to collect significantly more resources than competitors despite using the same limited hardware.

    Robotics & autonomySensors & measurement in products
    9 chapters

  • Over the Air · 16m

    Autonomous Navigation in GPS-Denied Environments with Ben Williams, COO at Exyn Technologies

    Ben Williams, COO of Exyn Technologies, explains how his company uses SLAM (simultaneous localization and mapping) to enable autonomous navigation in GPS-denied environments like mines and caves. He details the engineering challenge of operating in dusty, particulate-heavy conditions, describing how early systems would crash until they developed an action framework to detect and avoid these hazards. Williams highlights the transition from manual piloting to fully autonomous operations, noting partnerships with hardware providers like Boston Dynamics and software firms like Cesium to build a multimodal autonomy ecosystem.

    Embedded firmwareRobotics & autonomy
    6 chapters

  • Embedded.fm · 55m

    377: Robot at the Park

    Erin Kennedy, founder of Robot Missions and known as RobotGrrl, discusses her work building robots to clean plastic from shorelines and her involvement in the Hackaday Dream Team project to reduce ghost fishing gear. She details the technical challenges of her 'Bowie' robot, including navigating sand, using TensorFlow for debris detection, and the limitations of GPS versus AR markers for localization. Kennedy also shares insights on her unconventional learning path, her part-time role at an innovation lab, and her recent 'fog voxel' hobby project using piezo discs to create volumetric pixels.

    Robotics & autonomyAI & machine learning
    10 chapters

  • The Robot Report Podcast · 1h 20m

    Autonomous vehicle technology deep dive at John Deere and Zoox

    Jorge Heraud, VP of Autonomy at John Deere, explains how the company leverages its acquisition of Blue River Technology and Spark AI to deploy Level 4 autonomous tractors, using computer vision and machine learning to enable 'See and Spray' and autonomous tillage while managing rural connectivity challenges. RJ He and Ryan McMichael, technology leaders at Zoox, discuss their purpose-built, steering-wheel-free robotaxi recently deployed on public roads in California, detailing a perception stack that integrates cameras, LIDAR, radar, and thermal imaging to ensure safety through redundancy and diversity. The episode contrasts the agricultural application of autonomy, focused on labor relief and precision, with the urban mobility focus of Zoox, highlighting the technical hurdles of sensor fusion, edge cases, and the critical importance of building a robust safety case for public acceptance.

    Robotics & autonomySensors & measurement in products
    8 chapters

The summary, chapters and topics are generated from the episode’s audio by this directory, not written by Robot Talk. The episode itself belongs to its publisher.