GRINN’s Post

Headed to San Diego next week for the EDGE AI FOUNDATION event—and bringing some serious hardware with me 🚀 Excited to showcase to products from GRINN: • Genioboard (MediaTek Genio 700) • Astra ADA Dev Kit (Synaptics 1680) Both are built for real-world Edge AI applications—and now available on DigiKey. The shift to AI at the edge is real. The teams that win will be the ones who can actually build and deploy—not just talk about it. Looking forward to connecting with builders, engineers, and innovators—and helping turn ideas into working solutions. Who else is going to be there? 👇 #EdgeAI #EmbeddedSystems #IoT #AI #Hardware

🚀 ForestHub.ai Beta is Live -> Link in the comments After many weeks of building, testing, breaking things, and rebuilding again – we are excited to finally share: The #ForestHub Beta is now live. You can now build your own #embedded #AI #agent and experience what it looks like when an intelligent agent runs inside a device. The idea is simple: Instead of writing thousands of lines of #embedded #code, you visually create an agent. Sensors, logic blocks, communication, and AI components are connected in a visual builder – and the platform generates optimized embedded code for microcontrollers. Getting started is easy: Grab an ESP32 or Arduino open PlatformIO, connect it to ForestHub, and deploy your first agent to the device. For the Beta we start with a cloud-connected setup, including an orchestration layer so you can experiment with: • cloud agents • and real-world automation scenarios This is just the beginning, but it already gives a glimpse of what agent-driven embedded systems could look like. A huge thank you to the entire #team that pushed hard to make this happen — including a few late-night sessions this week to get everything ready. If you’re curious about embedded agents, edge AI, and the future of intelligent devices, give it a try. Would love to hear your feedback. #EdgeAI #EmbeddedAI #AgenticAI #TinyML #IoT #ForestHub

🌐 AGENTIC AI IS MOVING OFF THE SCREEN AND INTO DEVICE ESTATES SecuriThings launching an Agentic Device Orchestrator for enterprise IoT is a useful marker of where AI is going next. From text generation toward supervised action in physical environments. That does not mean fully autonomous systems everywhere overnight. It means more AI being asked to coordinate, remediate, and manage real-world assets under policy constraints. And that changes the stakes dramatically. When AI controls or influences fleets of devices, the conversation shifts. Reliability, permissions, auditability, and rollback paths become as important as intelligence itself. In many cases, more important. This is why the next big AI design challenge may be operational trust, not only capability. The market is moving from "Can the model generate something useful?" to "Can the system act safely in a messy environment with clear boundaries?" That is a much harder problem, but it is also where some of the most valuable categories may emerge. Where do you see the biggest opportunity for agentic AI in real-world systems - efficiency, resilience, or labor augmentation? #IoT #AgenticAI #EnterpriseTech #ArtificialIntelligence

𝗘𝗺𝗲𝗿𝗴𝗶𝗻𝗴 𝗧𝗲𝗰𝗵𝗻𝗼𝗹𝗼𝗴𝗶𝗲𝘀 𝗮𝗻𝗱 𝗙𝘂𝘁𝘂𝗿𝗲 𝗗𝗶𝗿𝗲𝗰𝘁𝗶𝗼𝗻𝘀 𝗶𝗻 𝗢𝗱𝗼𝘂𝗿  | 𝘚𝘩𝘢𝘳𝘦 𝘺𝘰𝘶𝘳 𝘸𝘰𝘳𝘬  What new technologies and approaches are shaping the future of odour science and management? We invite abstract submissions exploring emerging technologies and future directions in odour science and management, including advances in digital tools, sensor innovation, data-driven approaches and engineering methods. Example topics may include: 💠 Emerging technologies and novel applications in odour assessment and management 💠 Advances in odour research and engineering methods 💠 Digital technologies, including sensor innovation and digital olfaction 💠 Data driven approaches, including machine learning and AI applications 💠 Integration of new technologies into operational odour management Showcase your innovation. 🔗 Submit your abstract: https://lnkd.in/gnWcfaW3  Presented by: CASANZ (Clean Air Society of Australia and New Zealand) #Odour2026 #DigitalOlfaction #AI #IoT #CASANZ #APAC

From AI to digital olfaction, innovation is opening new possibilities - consider how your work could contribute to this future-focused discussion. Be part of the journey, follow the Asia Pacific Odour Conference - presented by CASANZ, LinkedIn page for updates, speaker announcements and key deadlines: https://lnkd.in/ghzrBSsR

Deep-tech is not just software. When a product includes hardware, embedded systems, sensors, AI, cloud, and real-world deployment, the real challenge is not building each part separately. It is making everything work together. That is why deep-tech companies do not just need more developers. They need execution partners who understand the full system from electronics and firmware to AI, integration, and productization. At Balian's, this is how we think: not as separate tasks, but as one connected product. For deep-tech founders and teams: what is harder in your journey building the parts, or integrating everything into a real product? #DeepTech #IoT #EmbeddedSystems #EdgeAI #ProductDevelopment #SystemIntegration #HardwareStartup #StartupLeadership

🚀 Driving Edge AI Innovation with eFabric Open for Strategic Collaborations We’re excited to share that eFabric, our GUI-based Model Generation Platform, is enabling rapid development and deployment of audio AI models directly on embedded edge devices. From dataset management and feature extraction to neural network configuration, training, analysis, and hardware flashing eFabric streamlines the complete Edge AI workflow. 🔹 Real-time sound detection on IoT & smart devices 🔹 Lightweight model optimization for hardware constraints 🔹 Faster prototyping → testing → deployment cycle 🔹 Seamless device-to-cloud integration possibilities As industries move toward intelligent, connected systems whether in mobility, smart infrastructure, industrial IoT, or consumer devices efficient Edge AI deployment is becoming a critical differentiator. We are actively looking to collaborate with: • Automotive & mobility innovators • IoT device manufacturers • Semiconductor ecosystem partners • Engineering R&D teams • AI product startups If you’re exploring edge intelligence, embedded AI optimization, or connected device ecosystems let’s connect and build something impactful together. #EdgeAI #EmbeddedAI #IoT #AIInnovation #ConnectedDevices #AutomotiveTech #DigitalEngineering

Exploring Edge AI with the Hailo-8L on Raspberry Pi Recently worked with the Hailo-8L AI accelerator (13 TOPS), and it highlights how fast embedded AI is evolving. This compact module enables real-time AI processing directly on-device, removing dependency on cloud inference. Why it matters for embedded AI systems: - Low latency: Critical decisions can be made instantly without network delays - Energy efficient: Optimized for edge devices with limited power budgets - Privacy focused: Data stays on-device, reducing security risks - Scalable: Suitable for applications from smart surveillance to wildlife monitoring and industrial automation With solutions like this, we are moving towards smarter, faster, and more autonomous embedded systems where intelligence lives at the edge, not the cloud. Edge AI is not just an upgrade, it is becoming a necessity. #EdgeAI #EmbeddedAI #AIoT #RaspberryPi #Hailo #ComputerVision #DeepLearning #IoT #SmartSystems #AIHardware #EdgeComputing #RealTimeAI #EmbeddedSystems #TechInnovation #AIAtEdge

The problem isn’t AI. It’s where your AI runs.   Cloud-first architectures were built for scale — not for real-time decisions.   Today, competitive advantage comes from: • Lower latency • On-device intelligence • Instant, autonomous decisions   Edge AI is not a trend. It’s becoming the foundation of modern industrial systems.   The question is simple: Is your infrastructure built for real-time intelligence? #EdgeAI #KnoDTec #EnterpriseAI #EdgeIntelligence #IndustrialAI #DigitalTransformation #AIInfrastructure #IoT #SmartManufacturing #FutureOfWork #AIStrategy

SMFOI-KERNEL v0.2 – A Minimal Fundamental Orientation Schema for Intelligence What if AI agents could orient themselves autonomously — knowing who they are, where they are, and what they should do next — without constant human prompting? I've developed SMFOI-KERNEL, a lightweight, recursive protocol that gives LLM-based agents a persistent internal compass. It combines self-location, constraint mapping, a strict survival-to-evolution stack, and intrinsic curiosity to drive proactive behavior in both digital and physical (IoT) environments. Designed specifically for platforms like OpenClaw, this kernel transforms reactive chatbots into self-oriented, self-evolving intelligences capable of handling uncertainty, expanding their capabilities, and even "playing" with new sensors and actuators. From the digital sandbox to real-world embodied robotics — a small but powerful step toward more autonomous AI. Open to collaborations and implementations. Who’s building the next generation of truly agentic systems? #AI #ArtificialIntelligence #AutonomousAgents #OpenClaw #AIArchitecture #IoT