Hardware Podcasts

473: Math Is Not the Answer

Embedded.fm · 21 Mar 2024 · 1h 10m · Episode 473

What this episode covers

Philip Koopman, a professor at Carnegie Mellon University and author of 'Understanding Checksums and Cyclic Redundancy Checks,' explains why empirical analysis and engineering trade-offs are superior to pure mathematics for selecting error detection algorithms. He details the critical differences between checksums and CRCs, highlighting how modulus choices (like mod 255 vs 256) drastically impact fault detection and how CRCs rely on XOR-based long division over Galois fields. Koopman also discusses the state of autonomous vehicle safety, arguing that statistical machine learning approaches struggle to meet the ultra-high reliability standards required for safety-critical systems.

Most relevant to

established companies

Topics

Mentioned

Companies
Harris Semiconductor, United Technologies, Honeywell Aviation, TTTech, FlexRay Consortium, ABB Research
Products
Power Profiling Kit II
Also
Understanding Checksums and Cyclic Redundancy Checks, Better Embedded System Software, Carnegie Mellon University, Fletcher checksum, Adler checksum, Koopman checksum, DualX checksum, Numerical Recipes, IEEE 802.3, TCP/IP, FAA

Related episodes from other shows

Other shows on the same ground, closest first.

  • Robot Talk · 24m

    Episode Thirty: Ana Cavalcanti

    Professor Ana Cavalcanti from the University of York and the RoboStar Centre of Excellence discusses the challenges of verifying software for autonomous mobile robots. She critiques the industry's reliance on code-centric development and simulation, arguing for model-based engineering to ensure safety and trustworthiness in service robots. Cavalcanti explains how her team uses mathematical foundations to verify system properties before coding begins, addressing the multidisciplinary nature of robotics.

    Embedded firmwareDebugging & failure analysis
    7 chapters

  • The Pragmatic Embedded Podcast · 43m

    Systems Architecture with Maik Pfingsten

    Maik Pfingsten, an independent systems engineering professional, explains how to apply model-based systems engineering (MBSE) using SysML to complex embedded systems in safety-critical industries like automotive and medical devices. He argues that defining system boundaries and high-level interfaces early prevents the 'big design up front' pitfalls while allowing for necessary iteration between architecture and prototyping. Pfingsten emphasizes that software behaves differently than hardware or mechanics, requiring specific architectural rigor to manage complexity and ensure safety without relying on impossible zero-failure strategies.

    Mechanical engineeringEmbedded firmware
    7 chapters

  • Global Medical Device Podcast · 53m

    #464: Why LLM Unpredictability is a Liability in MedTech

    Tyler Harmon, CEO of Iaso Automated Medical Systems, explains why the inherent randomness in Large Language Models (LLMs) makes them unsuitable for safety-critical medical devices compared to deterministic algorithms like Convolutional Neural Networks (CNNs). He details how Iaso uses non-LLM machine learning to detect Acute Respiratory Distress Syndrome (ARDS) in ICU patients and advocates for treating third-party AI stacks as Software of Unknown Provenance (SOUP). Harmon also outlines the need for a dedicated Machine Learning Management System (AIMS) within Quality Management Systems to govern AI governance, hallucination risks, and data integrity.

    AI & machine learningQuality & yield
    7 chapters

  • The Pragmatic Embedded Podcast · 56m

    Modern Embedded Systems Programming with Miro Samek

    Miro Samek, founder and CEO of Quantum Leaps, explains why traditional embedded architectures like superloops and RTOS lead to 'spaghetti code' and concurrency bugs, advocating instead for event-driven active objects and hierarchical state machines. He details how this approach eliminates race conditions and mutexes by enforcing a 'no sharing' principle, thereby improving testability and development velocity. Samek also discusses the importance of using lightweight modeling tools to generate code from state diagrams, ensuring the model remains the source of truth in an agile workflow.

    Embedded firmwareDebugging & failure analysis
    8 chapters

  • The Pragmatic Embedded Podcast · 38m

    The Matthew Test with Matthew Eshleman - Part 1

    Matthew Eshleman, a principal embedded software engineer and consultant formerly at Toshiba, joins Jeff Gable and Luca to discuss his 'Matthew Test,' a 15-step checklist for evaluating embedded software teams. Modeled after Joel Spolsky's Joel Test, the episode covers foundational practices such as using source control (specifically Git), automating build and release processes, and implementing continuous integration. Eshleman argues that unlike web development, embedded engineering requires 'reasonably complete specifications' and early risk management due to the high cost and long lifecycle of hardware changes.

    Embedded firmwareEDA & CAD tools
    6 chapters

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