Hardware Podcasts

Episode 76: Robots on the railways - Andrew Starr

Robot Talk · 8 Mar 2024 · 34m · Episode 76

What this episode covers

Andrew Starr, Professor of Maintenance Systems at Cranfield University, discusses the application of robotics in industrial maintenance and railway inspection. He details his early work at Ford monitoring car assembly robots for positioning errors and component degradation, highlighting the limitations of pure data-driven AI in predicting physical failures like corrosion. Starr then describes the development of autonomous railway inspection robots, from small-scale Arduino models to a full-sized Clearpath Warthog UGV and a retrofitted proprietary track trolley, emphasizing the challenges of GPS denial in tunnels and the necessity of data fusion.

Most relevant to

established companies

Topics

Mentioned

Companies
Cranfield University, Ford, Network Rail
Products
Clearpath Warthog
Chips
Arduino
Tools
LIDAR, RTK

Related episodes from other shows

Other shows on the same ground, closest first.

  • The Amp Hour · 1h 16m

    An Interview with Chris Osterwood

    Chris Osterwood, founder of Capable Robot Components, discusses his journey from Carnegie Mellon University to serving as CTO at Carnegie Robotics, where he worked on autonomous sewer inspection robots and agricultural systems. He explains the specific engineering challenges of ground robotics, such as managing weight, power, and harsh environments like sewers and cornfields. Osterwood details his new venture, which aims to provide reliable, embeddable infrastructure components like gigabit Ethernet switches and diagnostic USB hubs to solve the 'make vs. buy' dilemma for robot builders.

    Starting a product companyMechanical engineering
    8 chapters

  • The Next Byte Podcast · 15m

    196. Saving America’s Collapsing Infrastructure

    Hosts Daniel and Forbode discuss research from EPFL on using AI to detect cracks in America's aging infrastructure, specifically focusing on railroads and bridges. They explain how a new 'explainable AI' method reduces labeling effort by 90% compared to traditional pixel-by-pixel classification, allowing for continuous monitoring via instrumented rail cars or potentially buses. The hosts highlight the shift from subjective, costly manual inspections to a complementary model where automation flags issues for human experts to verify.

    AI & machine learningEnergy & climate tech
    5 chapters

  • The Robot Report Podcast · 1h 05m

    Transforming Asset Management with Physical AI

    Christian Pedersen, Chief Product Officer at IFS, explains how the enterprise software company is integrating physical AI and robotics into its asset management platform. He discusses the shift from digital-only AI to 'physical AI,' where robots like Boston Dynamics' Spot and Unitree's Neo act as workers alongside humans and digital agents. Pedersen highlights the need for specialized robotic skills, the importance of sensor data for predictive maintenance, and the challenge of ensuring AI accuracy in industrial settings to prevent costly errors.

    Connected products & IoTAI & machine learning
    8 chapters

  • The Amp Hour · 1h 19m

    An Interview with Gerry Roston

    Gerry Roston, CEO of Savionics and Executive in Residence at TechTown Detroit, discusses his career from JPL robotics to industrial IoT. He details building the 'Robbie' prototype for the Mars Sojourner rover, using PDP-11 and VAX computers, and later autonomous vehicles at CMU. At Savionics, he explains deploying wireless sensor nodes to monitor stamping machines, avoiding PLC integration to mitigate security risks, and using customer discovery to pivot from energy monitoring to predictive maintenance.

    Embedded firmwareSensors & measurement in products
    7 chapters

  • Manufacturing Happy Hour · 48m

    181: Leadership and Ethics in the Adoption of Robotics with Ryan Gariepy, CTO of Clearpath Robotics & OTTO Motors

    Ryan Gariepy, CTO of Clearpath Robotics and OTTO Motors (both acquired by Rockwell Automation), discusses the evolution of his companies from garage startups to enterprise solutions. He explains how observing the stark contrast between legacy industrial automation and Kiva Systems' technology revealed the future of robotics, leading to the creation of Clearpath for R&D platforms and OTTO for material handling. Gariepy emphasizes that successful robotics adoption requires balancing hype with signal, focusing on specific high-value applications, and prioritizing human safety and ethical deployment in both manufacturing and broader societal contexts.

    Leadership, teams & cultureRobotics & autonomy
    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.