Twinning action for spreading excellence in Artificial Intelligence of Things

A new AIoTwin blog post is now available, presenting recent research on running WebAssembly (WASM) on resource-constrained IoT devices. The article explores the practical limits, trade-offs, and opportunities of using WASM on microcontrollers with kilobytes of RAM, highlighting how lightweight runtimes can provide portability and sandboxed execution while balancing energy, memory, and performance constraints. Key topics covered in the blog post include: Benchmarking two lightweight WASM runtimes (wasm3 and WAMR) on three microcontrollers: Raspberry Pi Pico, ESP32-C6, and Nordic nRF5340 Measuring execution time, memory footprint, and energy consumption Comparing WASM execution with native C for typical IoT workloads (bubble sort and CRC-16) Trade-offs between performance, portability, and sandboxing in constrained devices Insights into runtime selection for modular, cross-platform IoT applications Authored by Mislav Has, the blogpost explains how WASM can simplify cross-platform deployment, enable secure execution of dynamic modules, and support unified toolchains, while highlighting scenarios where native code remains essential for performance-critical or ultra-low-power tasks. We are also pleased to announce that this work, titled “WebAssembly on Resource-Constrained IoT Devices: Performance, Efficiency, and Portability,” was presented at the ScaleSys workshop within the IoT 2025 conference . More details on the technical implementation, results, and evaluation scenarios can be found in the article here . Read the full blog post [ here ].

IoT applications are increasingly pushing the limits of cloud-only architectures. From autonomous driving to real-time monitoring, many services need ultra-low latency and can’t rely on distant cloud servers alone. In our latest AIoTwin blog post, Ivan Čilić breaks down how the Computing Continuum (CC) can meet these demands—and why adaptive, QoS-aware load balancing is essential for reliable performance. This work comes from his doctoral research at FER and ongoing collaboration with the  DSG Group at TU Wien as part of the AIoTwin project. Key highlights : How dynamic conditions in the CC break traditional load-balancing approaches Introducing QEdgeProxy — a decentralized, QoS-aware load balancer How QEdgeProxy uses local data to learn which service instances meet QoS requirements in real time How a Kubernetes-based implementation enables lightweight, scalable deployment Ivan’s work shows that decentralized QoS-aware load balancing is not only feasible but necessary to keep edge services responsive under real-world conditions. Read the full post here .

Cities today generate massive volumes of mobility data through sensors such as cameras, GPS, and roadside units. Traditional cloud-based systems struggle to process this data efficiently, particularly when real-time responsiveness is required. In a new AIoTwin blog post, Ivan Kralj explores how cloudlets , Edge AI , and adaptive learning are enabling scalable and accurate traffic prediction for smart cities. This work originates from Ivan’s research exchange in Sweden and ongoing collaboration with RISE Research Institutes of Sweden . Key Highlights: Traffic prediction is uniquely challenging due to the interconnected, dynamic nature of road networks. Fully centralized systems face limitations in scalability, latency, and reliability. Semi-decentralized training of Spatio-Temporal Graph Neural Networks (ST-GNNs) across distributed cloudlets improves resilience and efficiency. Sudden traffic events require specialized evaluation metrics — introducing Sudden Change in Speed Rate (SCSR) to detect congestion onset and recovery. Adaptive graph pruning reduces communication overhead while maintaining high prediction accuracy. Ivan’s research demonstrates that decentralized ST-GNN training is both practical and essential for future smart mobility systems, offering a more responsive and bandwidth-efficient approach to traffic prediction. Read the full post here .

A new AIoTwin blog post is now available, presenting recent research on how predictive edge AI can make smart agriculture more efficient, scalable, and resilient. The article explores how lightweight AI models, that can run on edge devices, can drastically reduce data transmission while maintaining high accuracy. This approach enables smarter energy use, improved resilience in low-connectivity areas, and scalable deployment even on farms without historical sensor data. Key topics covered in the blog post include: Predictive edge AI that transmits data only when necessary In-situ, cross-site, and satellite-to-ground deployment scenarios Over 92% reduction in data transmission at a 1.0 °C threshold Results from agrometeorological stations across Croatia How this supports climate-smart, data-efficient agriculture Authored by Dora Kreković , following her short-term exchange at Technische Universität Berlin , the blog post explains how the approach performs across three deployment scenarios: in-situ training and inference, cross-site model transfer, and satellite-trained models for farms without historical data. The results demonstrate how predictive edge AI can support climate-smart and data-efficient agriculture, even in regions with limited sensing infrastructure. We are also pleased to announce that this work has been accepted for presentation at the AIOT Conference 2025 , taking place in Osaka, Japan, from 3 to 5 December . Read the full blog post here . More updates and technical articles from AIoTwin researchers will follow soon.

A new article titled “Handling Dynamics in HFL Pipelines: The Need for Adaptive Orchestration” , authored by Ana Petra Jukić , has been published as part of the ongoing AIoTwin project blog series . The article reflects work carried out during and following the author’s short-term exchange at Technische Universität Berlin , focused on advanced orchestration strategies for Hierarchical Federated Learning (HFL). The text discusses key challenges arising in dynamic environments, where HFL pipelines may become suboptimal due to node arrivals or departures, fluctuations in resource availability, or non-uniform data distributions. It outlines the importance of adaptive orchestration and provides an overview of the AIoTwin orchestration middleware , developed on Kubernetes, which monitors performance, resource usage, and cost metrics to support intelligent and timely reconfiguration of HFL pipelines. The article also presents the Reconfiguration Validation Algorithm (RVA) , a method used to assess whether newly applied configurations offer real improvements without exceeding defined budget constraints. The full article is available here .

The AIoTwin project has released a new article in its ongoing blog series, focusing on Hierarchical Federated Learning (HFL) - an emerging approach designed to enhance the scalability and efficiency of distributed learning across Cloud-Edge-IoT (CEI) environments. While traditional Federated Learning (FL) enables collaborative model training without sharing raw data, it often faces limitations in scalability and communication efficiency. HFL addresses these challenges by introducing an intermediate layer of edge aggregators , creating a robust, multi-tier learning architecture that supports more decentralized and resilient AI applications. The blog post, authored by Katarina Vuknić following her research exchange at the Distributed Systems Group (DSG) , TU Wien , as part of the AIoTwin project , explores several key aspects of HFL: Communication Efficiency: How HFL significantly reduces uplink communication costs compared to flat FL architectures. Trade-offs: Analyzing challenges such as convergence delay, gradient variance, and model divergence in multi-level aggregation setups. Real-World Impact: Demonstrating the use of HFL in applications like smart farming and intelligent traffic management systems. The AIoTwin Solution: Presenting the open-source Extension of the Flower Framework for HFL , a Python-based component that enables scalable client, local, and global aggregation across the CEI continuum. The AIoTwin HFL solution offers a flexible, modular implementation of federated learning services designed to support distributed AI orchestration from the edge to the cloud. It represents a key step forward in advancing open, adaptive, and privacy-preserving machine learning for next-generation intelligent systems. The full article is available here .

A new blog post has been published focusing on the Artificial Intelligence of Things (AIoT) and the outcomes of the AIoTwin project. AIoT is revolutionizing the way humans and systems interact with the physical world by embedding intelligence directly into devices — from city cameras to industrial sensors. By integrating AI across Cloud-Edge-IoT (CEI) environments, AIoT enables systems that can learn, reason, and make autonomous decisions in real time. However, this transition from cloud-only or edge-only AI to a full computing continuum introduces new challenges, including determining where AI models should run — in the cloud, at the edge, or across both. The first blog post, authored by Ivana Podnar Zarko , explores several key aspects: Real-time traffic monitoring as a practical AIoT use case Cloud vs. edge-based AI: accuracy, latency, network load, and privacy trade-offs Why hybrid deployments across the computing continuum are the future Key technical challenges: optimal service placement, dynamic routing, federated learning, fault tolerance, and model updates As a solution, the AIoTwin Orchestration Middleware has been introduced. Developed by researchers at IoTLab@FER (Ivan Čilić, Ana Petra Jukić, Katarina Vuknić) in collaboration with researchers from Technische Universität Wien (Pantelis Frangoudis, Anna Lackinger, Ilir Murturi, Alireza Furutanpey, Schahram Dustdar), the middleware delivers a suite of open-source tools — including the fl-orchestrator , Flower extension for HFL , and QEdgeProxy . These tools form the foundation for scalable, adaptive, and intelligent AIoT deployments spanning from the edge to the cloud. The full blog post is available here , with additional technical articles from consortium members to follow.

Berlin, 24.6.2025.  The AIoTwin consortium is on the road again shortly after the 3rd Summer School. Fraunhofer FOKUS , one of Germany’s leading institutes for applied research in ICT, organized the Business Development & Innovation Transfer Workshop to foster the commercialization of Artificial Intelligence of Things (AIoT) innovations developed within the AIoTwin project . A full-day focused workshop placed special emphasis on exploring the business potential of the consortium’s AIoT middleware, offering insights into technology transfer, market readiness, and commercialization strategies. The one-day workshop featured a series of targeted presentations and collaborative challenges, fostering discussion and practical reflection on business development and innovation transfer strategies. Topics included: Lean Canvas for structuring and validating AIoT business ideas Technology Readiness Level (TRL), Market Readiness (MRL), and Product Readiness (PRL) assessments Business model design and approaches for real-world implementation Participants gained a deeper understanding of how to translate AIoT research into viable products and services. They left with validated business concepts, clearer market positioning, and concrete next steps for implementation. For more pictures from the workshop, open the detailed news content.

The detailed program for the third and final AIoTwin Summer School has been released. Set to take place from June 16 to June 18, 2025 , in Split, Croatia , the event will focus on the intersection of Internet of Things (IoT) and Artificial Intelligence (AI) . The Summer School will be co-located with the 10th SpliTech Conference , offering attendees the unique opportunity to participate in both events. This collaboration will provide enhanced networking and learning opportunities, making it a must-attend for those interested in the latest developments in IoT and AI. The detailed programme is available  here . More information about the summer school is available here .

The AIoTwin consortium met in Vienna from May 6–9, 2025 for the plenary meeting, where they reviewed the project's progress and discussed the next steps. Alongside the plenary sessions, participants attended a talk, a tutorial, and a workshop, each aimed at enhancing their knowledge and supporting collaboration. The meeting kicked off with a dedicated PhD student roundtable , fostering dialogue and exchange among PhD candidates within the consortium. This session provided a platform for emerging researchers to discuss their work and engage with fellow researchers. On the second day, participants took part in an Innovation Management Workshop organized by TU Wien's Innovation Incubation Center (I2C) , with Alexandra Negoescu leading the session. The workshop concentrated on strategies for translating academic research into practical applications, offering valuable insights into innovation, entrepreneurship, and technology transfer. Later in the day, Prof. Schahram Dustdar delivered a talk on “ The Art and Science of Writing Publishable Research ”, offering guidance on navigating the paper publishing process. The day concluded with a tutorial on " Autonomous Orchestration of Computing Continuum Systems through Active Inference ", presented by Boris Sedlak (TUW) . The session provided practical knowledge, enhancing participants’ understanding of these emerging technologies. The final two days were dedicated to plenary sessions, during which consortium members presented updates on work packages, shared progress, and engaged in discussions. The consortium will be meeting soon again, in Split, Croatia, for the third and final AIoTwin Summer School , co-located with the SpliTech Conference. Find out how to participate here .   For more pictures from the plenary meeting, open the detailed news content.