ADAGOS’ Post

You can still register for our next #tinyMLTalks Webcast: High-Efficiency Embedded Computer Vision on May 28th at 8 am PT with Sandeep M. Mistry, Principal SW Engineer and Developer Evangelist IoT at Arm and Alex Chang, Senior Product Manager, Himax Technologies, Inc.  https://lnkd.in/gkWqaxSf TinyML models have traditionally focused on gesture recognition, keyword spotting, and person detection. But what’s next? Learn how next-generation MCU silicon with an integrated energy-efficient NPU enables advanced computer vision models to be deployed to endpoints. We demonstrate face landmark detection and pose detection models running on a low-power microcontroller. #tinyAI #tinyML #ml #ai #machinelearning #artificialintelligence Evgeni Gousev Pete Bernard Olga Goremichina

#NDTArticle Alessandro Sardellitti’s latest thesis offers a view into how Industry 4.0 is reshaping quality control through advanced Non-Destructive Testing and Evaluation (NDT&E) methods. By integrating IoT sensors, AI, and real-time data analytics, this research paves the way for early defect detection and continuous, on-line inspections. Key highlights include the use of low-cost sensors for real-time measurements, optimized eddy current methods for thickness estimation and the application of dimensional analysis to simplify complex NDT&E challenges. Read the full research and its implications for Industry 4.0 here: https://lnkd.in/dX8Eyj9q ReJNDT Special Issue 'Frontiers in Nondestructive Testing and Evaluation: Academic Insights (2022/23)' captures the contents of outstanding master's and doctoral theses in the field of nondestructive testing. It is edited by Dr. Thomas Würschig, Roberto Henry Herrera, Cesar Camerini, aldo Canova, Michele Carboni, Younho Cho, Elena Jasiuniene & Antonello Tamburrino. #NDT #NonDestructiveTesting #Industry40 #NDE40 #Innovation #OpenAccess #ResearchInsights

Elevating computer vision with ihOS board: Real-time human pose detection at 5.0 Trillions Operations Per Second(5-TOPS)🤯🔥, outperforming with effortless precision. 🚀👤. . . . . . #ihOS #Upeya #ihOSBoard #IoT #AI #AIML #Automation #Future #Futuristic #Tech #Technology #SmartOffice #AI #FuturisticOffice #Innovation #edgecomputing #posedetection #camera #iotinnovation #openai #computervision #numpy #opencv #npu #neuralnetworks #neuralprocessingunit #cpu #gpu #detection #homeautomation

🌟𝐀𝐄𝐎𝐋𝐔𝐒: 𝐀𝐧 𝐀𝐟𝐟𝐨𝐫𝐝𝐚𝐛𝐥𝐞, 𝐦𝐢𝐧𝐢𝐚𝐭𝐮𝐫𝐢𝐬𝐄𝐝, 𝐜𝐥𝐎𝐮𝐝-𝐜𝐨𝐧𝐧𝐞𝐜𝐭𝐞𝐝 𝐬𝐲𝐬𝐭𝐞𝐦 𝐩𝐨𝐰𝐞𝐫𝐞𝐝 𝐛𝐲 𝐝𝐞𝐞𝐩 𝐋𝐞𝐚𝐫𝐧𝐢𝐧𝐠 𝐚𝐥𝐠𝐨𝐫𝐢𝐭𝐡𝐦𝐬 𝐟𝐨𝐫 𝐜𝐨𝐦𝐩𝐫𝐞𝐡𝐞𝐧𝐬𝐢𝐯𝐞 𝐚𝐢𝐫 𝐪𝐔𝐚𝐥𝐢𝐭𝐲 𝐦𝐞𝐚𝐒𝐮𝐫𝐞𝐦𝐞𝐧𝐭𝐬 𝐛𝐚𝐬𝐞𝐝 𝐨𝐧 𝐡𝐢𝐠𝐡𝐥𝐲 𝐢𝐧𝐭𝐞𝐠𝐫𝐚𝐭𝐞𝐝 𝐦𝐢𝐝-𝐈𝐑 𝐩𝐡𝐨𝐭𝐨𝐧𝐢𝐜  🌟 👉 Driven by an ambitious vision, AEOLUS Project is focused on creating an affordable, cutting-edge multi-gas photonic sensor and a cloud-connected, Big Data-driven smart sensing platform. Specifically, AEOLUS will: 🔹Capitalise on well-established Silicon (Si) platform, develop low cost and miniaturized, allowing for high integration sensing elements with enhanced performance 🔹Leverage CMOS compatible Germanium on Insulator processes and extend detection range up to ~10 μm 🔹Use wafer level processes to considerable minimize the sensor’s cost and footprint 🔹Demonstrate a system on chip integrated photonic sensor for multiple gases 🔹Use well established embedding PCB technologies to ensure proper thermal management 🔹Develop and validate Deep learning models that will provide emerging patterns, accurate chemometric analysis and predictions 🔹Leverage IoT testbed and demonstrate supporting IoT services, including air quality visualisation, real time health and safety security alerts, real time commands execution, and gamification 🔹Validate and Demonstrate affordable AEOLUS multi-gas smart sensing system to TRL7, and propel its comprising technologies beyond the technological ‘valley of death’ 🔹Deliver a holistic roadmap and business plan analysis for the path to market of sensing platform and exploitation plan Click to find out more 🔗 https://lnkd.in/dHYbbRFr #Photonics #Innovation Photonics21 #ai #InternetOfThings #PhotonicsEU #photonics

RS is proud to announce a new global franchise partnership with Synaptics Incorporated, a leader in AI-driven semiconductor solutions, aimed at transforming industrial applications with advanced edge AI and IoT technologies. We’re also excited to introduce the ASTRA development kit, which is engineered to accelerate edge AI processing for industrial environments. The kit is powered by Synaptics AI System-on-Chips (SoCs) and combines multi-core AI processors, NPU accelerators, and energy-efficient Edge AI inferencing capabilities. This enables real-time processing for machine learning, computer vision, and voice recognition - all critical for industrial use cases like control systems, smart automation, computer vision in manufacturing, and sensor-based monitoring. By deploying AI directly at the edge, the ASTRA kit is ideal for industrial IoT, autonomous systems, and smart factory applications. Join RS & Synaptics Incorporated as we pave the way for smarter, more efficient industrial systems with AI at the edge. https://lnkd.in/ezbFHeam #AI #EdgeAI #IndustrialIoT #MachineLearning #SmartAutomation #AIHardware #Semiconductors #Industry40 #partnership #demandcreation #designspark

🚀 Exciting News in Robotic Grasping Technologies!    I'm delighted to share that our paper, "Efficient Event-Based Robotic Grasping Perception using Hyperdimensional Computing," has been published in the prestigious Internet of Things (IoT) Journal by Elsevier, which ranks in the top 4% of journals in the field. This research introduces GraspHD, an innovative algorithm that significantly enhances the efficiency and effectiveness of robotic grasping by exploiting a Dynamic Vision sensor (DAVIS). By leveraging brain-inspired hyperdimensional computing (HDC), GraspHD operates at speeds 10x faster and offers an energy efficiency 26x higher than existing algorithms, maintaining robust performance in noisy environments. This work is a collaborative effort with esteemed authors Eman Hassan, Zhuowen Zou, Hanning Chen, Mohsen Imani, Yahya Zweiri, Hani Saleh, and Baker M..  Dive into our findings and explore the future of robotic applications!   🔗 Link to the paper: https://lnkd.in/dSy3h7eJ #Robotics #AI #Innovation #Research #Engineering #IoT #HyperdimensionalComputing

Dear LinkedIn Connections, I am thrilled to share that our latest research paper, "Augmented Multi-Agent DRL for Multi-Incentive Task Prioritization in Vehicular Crowdsensing" has been accepted for publication in the IEEE Internet of Things Journal and is now available in Early Access. This paper explores the emerging field of augmented intelligence, where we introduce a framework that merges human-like decision-making with AI-driven processes. Our framework, IMARL, integrates both human-centric and AI-based decision-making to optimize task prioritization in vehicular crowdsensing (VCS) systems within the social Internet of Vehicles (IoV). By leveraging augmented intelligence, IMARL enhances energy efficiency, reduces processing times, and adapts effectively to dynamic traffic conditions, offering significant advancements for VCS applications. Access the full paper here: https://lnkd.in/dKccm72x I want to express my deep gratitude to my co-authors, Bishmita H., Prof. Keshav Singh, Prof. Wan-Jen Huang, and Prof. Chih-Peng Li, for their invaluable guidance, support, and contributions to this work. #IOT #AI #DRL #IEEE #journal #crowdsensing #edgecomputing #vehicularnetworks #connectedvehicles #reinforcementlearning #augmentedintelligence

🌟 The publication of our latest paper, "NeuroCluster: Revolutionizing Energy Efficiency in Industrial IoT Networks," in the prestigious 2024 Second International Conference on Emerging Trends in Information Technology and Engineering (ICETITE). 📝 In this groundbreaking work, we introduce NeuroCluster, a cutting-edge approach that harnesses the power of neural networks to optimize energy usage and extend the lifetime of Industrial Internet of Things (IIoT) networks. 💡 🔍 Key Highlights: * Dynamic clustering strategies based on real-time energy consumption patterns and environmental factors. * Adaptive allocation of devices into clusters considering current energy states and predictive models of future energy demands. * Neural network architecture designed to capture intricate relationships between device characteristics, contextual variables, and energy consumption profiles. * Extensive simulations demonstrating significant improvements in energy efficiency compared to traditional clustering methods. * Robustness in diverse operational conditions, making it a scalable and adaptable solution for a wide range of IIoT applications. Our research represents a significant advancement in the field of IIoT, offering tangible solutions to address critical concerns surrounding energy efficiency in clustered environments. 🌐 Read the full paper and explore the future of IIoT energy optimization: [NeuroCluster: Revolutionizing Energy Efficiency in Industrial IoT Networks](https://lnkd.in/gBhRvuzj) 📚 #IIoT #EnergyEfficiency #NeuralNetworks #Innovation #ICETITE2024 #Engineering #Technology #IoT #IEEEConference 🚀

Explore what's possible at the IoT Edge with the #Astra #Machina AI DevKit, available soon from the Synaptics partner ecosystem!

🧠 Excited to share my latest innovation: The Neuromorphic Time-Series Compression (NTSC-BMP) Algorithm! Have you ever wondered how we could make machines process data more like the human brain? That's exactly what I set out to solve. 🔬 What is NTSC-BMP? 1- A brain-inspired compression algorithm that processes time-series data like human memory 2- Intelligently identifies and preserves critical data points 3- Efficiently summarizes routine information 4- Perfect for industrial IoT, sensor data, and real-time monitoring 🚀 Live Demo: I've implemented this algorithm using NASA's jet engine dataset as a proof of concept. You can try it yourself: https://lnkd.in/eUcK8TNF Key Features: 📊 Adjustable significance thresholds 🎯 Anomaly detection 💾 Multi-layer retention system 📈 Real-time visualization 🛠️ Use Cases: - Industrial sensor monitoring - Predictive maintenance - Time-series data analysis - IoT data processing 👨💻 Would love to connect with professionals interested in neuromorphic computing, data compression, or industrial IoT. Feel free to check out the live demo and share your thoughts! Department of Computer Science at NTU Nottingham Trent University Ahmad LOTFI #Innovation #DataScience #MachineLearning #Neuromorphic #Algorithm