967: AI for the Physical World, with Samsara's Praveen Murugesan - Super Data Science: ML & AI Podcast with Jon Krohn Recap
Podcast: Super Data Science: ML & AI Podcast with Jon Krohn
Published: 2026-02-17
Duration: 55 min
Summary
Praveen Murugesan discusses how Samsara leverages AI to enhance safety, efficiency, and sustainability in physical operations, processing an astonishing 20 trillion data points. The conversation highlights the challenges and innovations in running AI on the edge for real-world applications.
What Happened
In this episode, Jon Krohn welcomes Praveen Murugesan, VP of Engineering at Samsara, a leader in software solutions for physical systems. Murugesan shares insights from his experience overseeing the application of AI to interpret massive volumes of data—specifically 20 trillion data points covering 90 billion miles annually. He emphasizes the importance of data in creating impactful solutions that enhance safety and efficiency in sectors like transportation, construction, and logistics.
The discussion delves into the specific challenges of handling such vast data, with Murugesan noting that Samsara is likely collecting more real-world data than any other company, except perhaps Uber, where he previously worked. He describes Samsara's focus on safety and efficiency for businesses operating in low-margin environments, highlighting how they develop software and hardware solutions to address common operational challenges. Notably, the episode explores edge computing, particularly how AI models can run directly on devices like safety cameras installed in vehicles to monitor driver behavior and alert them to potential drowsiness.
Key Insights
- Samsara processes over 20 trillion data points annually, focusing on real-world applications.
- Safety and efficiency are critical themes for businesses in low-margin environments.
- Edge computing enables real-time AI inference, enhancing operational safety.
- Samsara's innovations include proprietary AI models tailored for various physical systems.
Key Questions Answered
What is Samsara's unique role in data processing for physical operations?
Samsara stands out by processing an enormous volume of data—20 trillion data points—focused on enhancing safety and efficiency in physical operations. Murugesan notes that they cater to various sectors such as construction, transportation, and logistics, all of which face common challenges related to safety, efficiency, and sustainability.
How does edge computing work in Samsara's applications?
Samsara employs edge computing to enable AI models to run directly on devices like safety cameras in trucks. This allows for real-time monitoring of driver behavior, such as detecting drowsiness. The inference happens on the edge, which requires efficient processing power, and allows for quicker feedback to improve safety.
What are the main challenges of handling such vast data volumes?
Murugesan explains that managing vast volumes of data presents both exciting opportunities and significant challenges. The key is to develop effective AI applications that can interpret the data intelligently while ensuring the models run efficiently on constrained hardware.
What types of safety solutions does Samsara provide?
Samsara focuses on safety solutions that include hardware like cameras installed in vehicles. These cameras utilize AI to analyze driver behavior and provide immediate feedback, helping to prevent accidents caused by fatigue or distraction.
How does Murugesan's background influence his work at Samsara?
Murugesan's previous experience at Uber shaped his desire to tackle physical world problems, which he continues at Samsara. His background helps inform the development of innovative solutions that leverage the vast data collected to enhance operational efficiency and safety.