Embedded AI - Intelligence at the Deep Edge

David Such

“Intelligence at the Deep Edge” is a podcast exploring the fascinating intersection of embedded systems and artificial intelligence. Dive into the world of cutting-edge technology as we discuss how AI is revolutionizing edge devices, enabling smarter sensors, efficient machine learning models, and real-time decision-making at the edge. Discover more on Embedded AI (https://medium.com/embedded-ai) — our companion publication where we detail the ideas, projects, and breakthroughs featured on the podcast. Help support the podcast - https://www.buzzsprout.com/2429696/support

  1. Why Humanoid Robots Need Two Clocks

    3d ago

    Why Humanoid Robots Need Two Clocks

    Send us Fan Mail A useful general-purpose robot has to do two things that fight each other. It has to think slowly enough to understand "put away the groceries," and it has to move fast enough to keep a grip on the milk carton without crushing it. The part that understands is large and slow. The part that moves has to be small and fast. You cannot run both on the same clock. This episode looks at the design now shipping on real robots: Vision-Language-Action models that simply run two clocks at once. A slow brain that thinks a handful of times a second, a fast brain that moves two hundred times a second, and a single note of intent passed between them. We walk through how Figure's Helix splits a 7-billion-parameter planner from an 80-million-parameter controller, why "action chunking" keeps the motion smooth, and how a March 2026 project squeezed the whole pipeline onto a 40-watt module with no cloud connection at all. This two-speed design is the same answer evolution reached, with the cortex deciding the goal and the cerebellum handling the reflexes. When biology and engineering independently land on the same structure, it is probably telling us something fundamental about what it takes to be intelligent inside a moving body.  Support the show If you are interested in learning more then please subscribe to the podcast or head over to https://medium.com/@reefwing, where there is lots more content on AI, IoT, robotics, drones, and development. To support us in bringing you this material, you can buy me a coffee or just provide feedback. We love feedback!

    23 min
  2. Who is Liable for Onboard AI?

    May 31

    Who is Liable for Onboard AI?

    Send us Fan Mail As foundation models move from the cloud into physical robots, a fundamental question emerges: who is accountable when an AI-controlled machine makes a decision that causes harm? In this episode, we examine the growing collision between embodied AI, functional safety, and emerging regulation. We explore how new frameworks such as the EU AI Act and the Machinery Regulation are reshaping expectations for developers, manufacturers, and deployers of intelligent robots. From humanoid robots and autonomous mobile manipulators to AI-enabled industrial machinery, the challenge is no longer simply making robots smarter. It is making them governable. We investigate a proposed architectural solution that is gaining traction across industry and academia: the hardware-isolated safety supervisor. By separating non-deterministic AI reasoning from deterministic safety-critical control systems, this approach aims to create clear lines of accountability while preserving the benefits of onboard intelligence. Along the way, we examine NVIDIA’s Cosmos Reason 2 model, the EmbodiedGovBench governance framework, emerging standards efforts, and the practical realities of deploying foundation models on embedded platforms. We also ask whether traditional functional safety concepts such as SIL and ASIL can adequately address the unique challenges posed by robots whose actions are selected by large vision-language models. The broader question is one that every roboticist, embedded engineer, and AI practitioner will soon face: when intelligence becomes local, autonomous, and physically embodied, what mechanisms ensure that accountability remains local too? Support the show If you are interested in learning more then please subscribe to the podcast or head over to https://medium.com/@reefwing, where there is lots more content on AI, IoT, robotics, drones, and development. To support us in bringing you this material, you can buy me a coffee or just provide feedback. We love feedback!

    24 min
  3. Squeezing AI into your Pocket

    May 28

    Squeezing AI into your Pocket

    Send us Fan Mail By 2026, language models have moved off the cloud and onto the device in your pocket. What was a research demonstration two years ago is now a routine engineering capability, and the centre of gravity for artificial intelligence has begun to migrate from distant data centres to local silicon. The episode traces the four engineering moves that made this possible. Quantization, which shrinks a model by storing its parameters with less precision. Optimized key-value caches, which let a model hold a long conversation without exhausting memory. Neural Processing Units, the dedicated AI accelerators now standard in flagship phones. And specialized frameworks such as LiteRT-LM and llama.cpp, which finally make all three usable from a single application. The consequences reach further than performance figures. Privacy becomes the default rather than a feature, because data never leaves the device. The cost structure of AI applications changes, because there are no per-query cloud fees. And the link between training capital and deployment capability begins to decouple, opening the door for small teams to ship genuine intelligence on hardware they already control. Support the show If you are interested in learning more then please subscribe to the podcast or head over to https://medium.com/@reefwing, where there is lots more content on AI, IoT, robotics, drones, and development. To support us in bringing you this material, you can buy me a coffee or just provide feedback. We love feedback!

    20 min
  4. A chip that controls a balancing propeller on seven microwatts

    May 14

    A chip that controls a balancing propeller on seven microwatts

    Send us Fan Mail Every battery-powered device you own has a quiet energy hog in it that nobody talks about. It is not the processor, it is not the radio, and it is not the screen. It is the analog-to-digital converter, the small piece of circuitry that translates the messy real world into the clean ones and zeros a computer can think about. For thirty years it has been the line item that decides how long your hearing aid, your pacemaker, or your soil sensor lasts on a battery. In March 2026, a team at the University of Michigan published a result that quietly removes that converter from the picture for a specific class of problems. Their bismuth selenide memristor runs a closed-loop control task at about seven microwatts, roughly a millionth of what a household LED bulb pulls. The chip does not run code in any conventional sense. The physics does the arithmetic, and the answer drives the motor directly. In this episode, we walk through what the device actually is, why removing the converter changes the energy budget by orders of magnitude, and which products land first when microwatt-class intelligence becomes buildable. We talk about hearing aids, implants, environmental sensors, and the small drones that have been waiting for this kind of result for a decade. We also talk about what this chip cannot do, because the press releases tend to skip that part. It will not run a language model. It will not recognise your face. It will run the reflexes underneath all of that, and the case for why those reflexes matter more than the cortex gets credit for is the through-line of the episode. Support the show If you are interested in learning more then please subscribe to the podcast or head over to https://medium.com/@reefwing, where there is lots more content on AI, IoT, robotics, drones, and development. To support us in bringing you this material, you can buy me a coffee or just provide feedback. We love feedback!

    15 min
  5. Why 95% of AI Deployments Fail

    May 7

    Why 95% of AI Deployments Fail

    Send us Fan Mail MIT's August 2025 study of 300 enterprise generative AI deployments found that 95% produced no measurable P&L impact. Gartner forecasts that more than 40% of agentic AI projects will be cancelled by 2027. McKinsey's State of AI 2025 identifies workflow redesign as the single strongest correlate with EBIT impact, yet only 21% of organisations have redesigned any workflows. The data converges on a structural conclusion: enterprise AI is failing because the operational substrate is inadequate, not because the models are. This episode examines the process-readiness gap, the misallocation pattern that concentrates investment in low-ROI front-office applications, and what the 5% of high performers do differently. It is an architectural argument, not a change-management one: AI is a linear amplifier acting on a pre-existing process, and the sign of the output depends on the sign of the input. Support the show If you are interested in learning more then please subscribe to the podcast or head over to https://medium.com/@reefwing, where there is lots more content on AI, IoT, robotics, drones, and development. To support us in bringing you this material, you can buy me a coffee or just provide feedback. We love feedback!

    23 min
  6. Why Humans and Robots must Dream

    Apr 25

    Why Humans and Robots must Dream

    Send us Fan Mail Put a blindfold on a sighted adult and the visual cortex starts being colonised by touch and hearing within forty-five minutes. Not weeks. Not days. Forty-five minutes. This is not a quirk of extreme cases. It is how the cortex works all the time. Every region of the brain is in continuous low-grade negotiation with its neighbours over territory, and the currency of that negotiation is activity. Stop using a subsystem and the neighbours move in, fast. This is the empirical foundation of a hypothesis from neuroscientist David Eagleman called the defensive activation theory: that REM sleep exists specifically to keep the visual cortex active during the eight hours each night when external input is unavailable, defending its territory against takeover by senses that never go offline. The theory itself is plausible but not yet directly proven. What is proven, and what matters more for engineers, is the underlying principle. A complex system with reconfigurable resources will silently lose capability in any subsystem that is not regularly exercised, even when nothing is actively trying to take that capability away. This is not catastrophic forgetting in the usual machine learning sense, where new training overwrites old parameters. This is something subtler and arguably more dangerous: passive territorial loss in any system that supports continuous adaptation. It shows up wherever capabilities are not being exercised in long-running adaptive AI: rarely-routed experts in mixture-of-experts models, underused sensor pipelines in multi-modal robotics, capabilities that drift out of online reinforcement learning agents over months of deployment. Most current architectures treat their structure as fixed by design. Biology treats its structure as continuously contested. This episode looks at what defensive activation reveals about a missing primitive in modern AI architecture. Current systems have two fundamental modes, training and inference. Brains have at least three, and the third one, the maintenance mode that operates during REM sleep, has no clean equivalent in the systems we build. We examine what this mode is doing structurally, why generative replay in continual learning is mechanistically closer to dreaming than the field usually acknowledges, and what a telemetry-driven maintenance subsystem might look like for embedded and edge AI. The closing argument is straightforward: if biology has been running this experiment for a few hundred million years and converged on internally-driven activation as the way to maintain a plastic computational substrate, the absence of an equivalent mechanism in our architectures is not a neutral design choice. It is a gap. Support the show If you are interested in learning more then please subscribe to the podcast or head over to https://medium.com/@reefwing, where there is lots more content on AI, IoT, robotics, drones, and development. To support us in bringing you this material, you can buy me a coffee or just provide feedback. We love feedback!

    24 min
  7. Sovereign AI and the End of the Borderless Cloud

    Apr 19

    Sovereign AI and the End of the Borderless Cloud

    Send us Fan Mail The borderless cloud era is ending. In the second week of January 2026, four government decisions announced in rapid succession made that shift undeniable: the UK activated its £500 million Sovereign AI Unit, France committed €109 billion, the UAE consolidated a $40 billion data centre portfolio, and the Trump administration revised chip export rules to China. In this episode, we examine why AI infrastructure is now being treated as a strategic national utility on par with energy and water, and what that means for engineers and boards making architectural decisions today. We map the global sovereign AI landscape, roughly 130 national initiatives across more than 50 countries, and separate political rhetoric from engineering reality. We examine the distinction between regulatory sovereignty (the legal authority to govern AI) and compute sovereignty (the physical capacity to run it), and explain why most nations have the first without the second. We cover China's full-stack response through Huawei's Ascend and CloudMatrix programme, a deliberate trade-off of efficiency for independence that is becoming a template other regions may follow. We draw on the Clipper chip precedent from the 1990s to show why embedded enforcement mechanisms in silicon create durable market incentives that are difficult to reverse. Support the show If you are interested in learning more then please subscribe to the podcast or head over to https://medium.com/@reefwing, where there is lots more content on AI, IoT, robotics, drones, and development. To support us in bringing you this material, you can buy me a coffee or just provide feedback. We love feedback!

    20 min
  8. The Agentic AI Reckoning: Autonomy, Safety, and the Edge

    Apr 13

    The Agentic AI Reckoning: Autonomy, Safety, and the Edge

    Send us Fan Mail In Q1 2026 the agentic AI conversation moved from theory to forensics. A crafted PDF triggered physical pump activation through a Claude MCP integration at an industrial facility, after an engineer used the same agent for routine document summarisation and SCADA writes. The hidden instructions used white-on-white text and base64 encoding, the agent treated the document content as instructions, and the legitimate credentials carried the action straight through to operational technology. The damage was physical. This episode walks through the Q1 2026 forensic record and asks the question the embedded community has been avoiding: what happens when an agent that rewrites its own action plan at runtime is wired to an actuator that does not have an undo button. Support the show If you are interested in learning more then please subscribe to the podcast or head over to https://medium.com/@reefwing, where there is lots more content on AI, IoT, robotics, drones, and development. To support us in bringing you this material, you can buy me a coffee or just provide feedback. We love feedback!

    25 min
See All (92)

About

“Intelligence at the Deep Edge” is a podcast exploring the fascinating intersection of embedded systems and artificial intelligence. Dive into the world of cutting-edge technology as we discuss how AI is revolutionizing edge devices, enabling smarter sensors, efficient machine learning models, and real-time decision-making at the edge. Discover more on Embedded AI (https://medium.com/embedded-ai) — our companion publication where we detail the ideas, projects, and breakthroughs featured on the podcast. Help support the podcast - https://www.buzzsprout.com/2429696/support

Information