Hardware Podcasts

TRI using generative AI to teach robots; Four Growers on developing harvesting robots

The Robot Report Podcast · 22 Sept 2023 · 1h 03m

AI in practice: 3 real uses of AI described

What this episode covers

Mike Oitsman reports from the FIRA Agricultural Robotics show, highlighting startups like Burrow, Trick Robotics, and Eggtonomy, while also noting Agility Robotics' new humanoid manufacturing facility. The episode features an interview with Brandon Contino, CEO and co-founder of Four Growers, who explains how his company uses a novel vacuum-based system to harvest tomatoes and cucumbers in greenhouses. Contino details the technical challenges of soft fruit handling, the strategic advantage of indoor environments for rapid iteration, and the company's expansion into AI-driven yield analytics and partnerships with seed giant Syngenta.

AI in practice

Where this conversation describes AI being used to build or deliver a product. Part of the AI in practice tracker.

Engineering & simulationFour Growers AI platformpicking faster than even a person can pick

The team used machine learning to build a custom data set and architecture that teaches the robot to understand plant concepts beyond physical characteristics, enabling it to determine ripeness and harvest tomatoes faster than a human.

“we've really created a data set and built our own architecture, really taught the AI to learn more than just physical characteristics of a plant, but to actually understand different concepts of that plant.”

Brandon Contino · Hear it at 33:58

Engineering & simulationTRI Large Behavior Modelstaught it 60 skills; hoping to have a 1000 skills by the end of next year

Toyota Research Institute used generative AI and diffusion models to learn dexterous manipulation skills from teleoperation demonstrations, allowing robots to acquire new skills in hours without hand-coding.

“TRI is doing is their AI policy is learning in the background, how to do that skill. They're claiming they can get these skills down in a matter of hours. … they've taught it 60 skills on their robots, they're hoping to have a 1000 skills by the end of next year.”

Steve Crowe · Hear it at 8:52

Engineering & simulationFour Growers AI platform1700000000 is spent in harvesting

The AI system analyzes real-time harvest data to predict yield levels (1-6 ripeness) per square meter, allowing growers to optimize logistics, scheduling, and pack house operations based on future availability.

“we're looking at all the tomatoes that are coming down the plant, and we're actually doing counts and providing back to the grower down to every square meter. Here are the number of kind of level 2 tomatoes you have … you can start to predict, you're harvesting and streamline the logistics.”

Brandon Contino · Hear it at 49:15

Most relevant to

early-stage companies and companies scaling up

Topics

Mentioned

Companies
Agility Robotics, Toyota Research Institute, Burrow, Trick Robotics, Eggtonomy, Four Growers, Syngenta Seeds, Carbon Robotics, FarmWise, Seed Spider, Verdant Robotics
Products
Digit

Related episodes from other shows

Other shows on the same ground, closest first.

  • Robot Talk · 28m

    Episode Seven - Agricultural Robotics: Farming for the future

    Dr Petra Bosilj of the University of Lincoln explains that robotic vision in agriculture is bottlenecked by a lack of annotated training data for complex tasks like distinguishing weeds from crops, though precision spraying can reduce herbicide use by 80%. Chris Chavasse, founder of Muddy Machines, describes his startup's development of lightweight, swarm-capable field robots for harvesting tender stem broccoli, designed to reduce soil compaction compared to heavy industrial machinery. The episode contrasts academic research into phenotyping and soil health with the commercial challenges of building rugged, soft-robotic end effectors and securing funding during the pandemic.

    Sensors & measurement in productsRobotics & autonomy
    8 chapters

  • The Next Byte Podcast · 27m

    164. Dyson’s Automated, Self-sustaining Strawberry Farm

    Daniel and Forbode discuss Dyson's transition from consumer electronics to agriculture, detailing their 36,000-acre automated farm in the UK. They highlight the integration of anaerobic digesters for energy, robotic strawberry pickers, and precision herbicide robots that reduce chemical use by 90%. The episode also covers Murata's soil sensor technology, which helps farmers optimize yields and resource usage through data-driven insights.

    Robotics & autonomyEnergy & climate tech
    8 chapters

  • How I Built This · 32m

    Doing the bees’ work with Thai Sade of BloomX

    Thai Sade, co-founder and CEO of BloomX, explains how his company uses robotics and AI to enhance crop pollination, addressing the decline of bee populations and the need for higher agricultural yields. He details the development of two main technologies: the Burrow, an autonomous vehicle that vibrates blueberry flowers to release pollen, and the Crosby, a manual electrostatic tool for transferring sticky pollen in crops like avocados. Sade discusses the company's validation process in Latin America, its $8 million in funding, and how increasing yield per hectare can help prevent rainforest deforestation.

    Robotics & autonomyAI & machine learning
    7 chapters

  • Robot Talk · 19m

    Episode Nineteen: Roopika Ravikanna

    Roopika Ravikanna, a PhD student at the University of Lincoln, discusses her research on autonomous agricultural robots, specifically the modular 'Thorvald' system developed with Saga Robotics. She explains how the robot assists farm workers in horticulture environments by navigating complex obstacles and avoiding soil compaction, while addressing human-robot interaction challenges like proximity and intuitive control. Ravikanna highlights that these cobots are designed to alleviate labor shortages rather than replace workers, with current deployments already underway at commercial sites like Clockhouse Farm.

    Robotics & autonomySensors & measurement in products
    7 chapters

  • The Next Byte Podcast · 26m

    44. How Computer Vision Is Changing Your World

    Hosts Daniel and Farbod discuss how computer vision AI is applied in manufacturing and agriculture, featuring insights from Plain Sight AI and a project by the Technical University of Denmark. They explain Plain Sight AI's three levels of vision systems for detecting hazardous objects on production lines, emphasizing that Level 3 instance segmentation offers superior sorting and counting capabilities. The hosts also detail a robot developed by PhD student Ronja Goldring to autonomously remove weeds from organic dairy farms using laser or electrocution, highlighting the importance of large, diverse training datasets and collaboration with end-users.

    AI & machine learningIndustrial & consumer product design
    5 chapters

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