Sovereign Edge AI Infrastructure — Scaleout

Research & Insights

Technical publications and perspectives from the Scaleout team.

Feb 16, 2026

Moving beyond detection to full autonomy at the edge

Autonomy is not a single AI model; it is a tightly coupled system of hardware-agnostic modules. Our stack transforms raw visual input into structured mission logic, enabling drones to maintain persistent target memory and execute complex maneuvers in communication-degraded environments.

Continue reading

Sigvard Dackevall

Machine Learning Engineer

Jan 10, 2026

Gradient inversion attacks in federated learning

We tested privacy risks against production-grade vision models like YOLO and ViT. In realistic settings, meaningful image reconstruction often collapses.

Viktor Valadi, Mattias Åkesson, et al.

Sep 4, 2025

Scaling Federated Learning to 10,000 asynchronous edge devices

Test your models against 10,000 drifting, dropping, and reconnecting clients. Scaleout Edge provides the modular infrastructure to move from lab prototypes to high-concurrency production deployments.

Scaleout

In-depth

Explorations of core concepts and systems behind Scaleout Edge, federated learning, and distributed AI.

Technical workshop

Hands-on workshop on building federated learning systems with Scaleout Edge.

To the workshop

Edge computing and AI

Edge AI: A Comprehensive Guide to Real-Time AI at the Edge

Explore the guide

What is federated learning?

An overview of how federated learning works and when it is useful

Learn about federated learning

Scalable federated machine learning

A deep dive into the paper explaning the foundation of Scaleut Edge

List of articles

Technical publications and perspectives from the Scaleout team.

Akkodis and Scaleout Accelerate Secure Edge AI: Nov 26, 2025; Akkodis and Scaleout are partnering to combine rugged industrial hardware with federated learning capabilities to accelerate the deployment of secure, scalable Edge AI solutions in mission-critical sectors like defense, energy, and transportation.
Unlocking Isolated Data Silos with Federated Self-Supervised Learning: Oct 9, 2025; Federated Learning enables collaboration without sharing medical data, but inconsistent annotations limit its impact. Self-Supervised Learning solves this by leveraging unannotated images to pre-train models, then fine-tuning on small labeled sets.
AI Everywhere points to Edge AI: Sep 29, 2025; AI everywhere requires low latency, bandwidth efficiency, and privacy. Federated Learning enables this by training models on-device, sharing only updates. This makes Edge AI secure, efficient, and adaptive.
Fighting Back Against Attacks in Federated Learning: Sep 22, 2025; Federated Learning enhances privacy but is vulnerable to data and model poisoning attacks. Tests with our new simulator show adaptive strategies like EE-Trimmed Mean are more resilient than traditional methods.
From Satellites to Fleets: Our Ongoing Research Initiatives: Sep 10, 2025; An overview of four current research projects at Scaleout, exploring advanced AI methods to tackle real-world challenges in satellite data processing, autonomous vehicles, and fleet intelligence.
Federated Learning with 10,000 Asynchronous Clients Using Scaleout Edge: Sep 4, 2025; A toolkit for large-scale federated learning that successfully simulated 10,000 asynchronous, intermittently connected clients, achieving stable model convergence while keeping data private on devices.
Data Selection on the Edge for Adaptive Federated Machine Learning: Sep 2, 2025; An edge-based data selection pipeline for federated machine learning, designed to improve model training without sending raw video streams to central servers, preserving privacy and reducing bandwidth.
Collaborative AI for Lung Cancer Detection: Federated Learning in Healthcare Without Sharing Patient Data: Aug 13, 2025; Federated learning allows hospitals to collaboratively train AI models for tasks like lung cancer detection without sharing sensitive patient data.
Fleet Intelligence with Mixture-of-Experts Federated Learning: Jul 4, 2025; A two-year project to develop a resource-efficient, privacy-preserving AI framework for connected fleets by combining Federated Learning and Mixture-of-Experts models.
Vertical Federated Learning with FEDn: May 16, 2025; Introduces Vertical Federated Learning in FEDn, allowing organizations with complementary data features to train a shared model collaboratively without sharing private data.
Exploring Python vs. C++ Clients - A Performance Deep Dive with FEDn Framework: Mar 17, 2025; Python vs C++ clients in FEDn: Python uses less idle memory but C++ is more efficient during training with better memory management and faster task completion. FEDn supports both simultaneously.
Scaleout secures new investment round to accelerate Cloud-Edge AI: Feb 26, 2025; Scaleout Systems secured 35 MSEK with Fairpoint Capital joining existing investors to enhance its FEDn framework for secure cloud-edge AI deployment across industrial IoT, automotive and defense sectors.
Distributed Continuous Machine Learning: Jan 28, 2025; Federated learning represents a shift in AI architecture where models travel to and train on distributed data sources, eliminating the need to centralize sensitive data.
Machine Learning in the Cloud-Edge Continuum: Dec 18, 2024; The future of AI is in the cloud-edge continuum, allowing organizations to use both centralized machine learning and edge computing with federated learning for privacy and resource optimization.
Scaleout Joins NATO's DIANA Programme to Advance Federated Intelligence in Conflict Zones: Dec 9, 2024; Scaleout's FEDAIR project for NATO's DIANA programme enables secure, decentralized ML model updates in conflict zones, allowing adaptation without compromising sensitive data.
Federated Learning Made Easy with FEDn: Dec 2, 2024; An article on using FEDn, a framework that enables federated learning by distributing training across multiple clients using PyTorch with FEDn's "compute package" system.
Guaranteeing Data Privacy for Clients in Federated Machine Learning: Nov 25, 2024; Federated Learning with Differential Privacy covers Gaussian noise for enhanced privacy and privacy measures in FL, addressing both record-level and client-level protections.
Federated Learning for Object Detection Using YOLO: Nov 5, 2024; YOLO object detection models integrated with federated learning enable privacy-preserving ML across distributed devices for autonomous vehicles, medical imaging, and manufacturing.
Hyperparameter tuning with Optuna and FEDn Python API: Sep 13, 2024; Using Optuna to optimize the server-side learning rate of FedAdam in FEDn, covering defining an objective function, tuning hyperparameters, and automating the process.
Enhancing Semiconductor Component Placement with Federated Learning: Aug 20, 2024; Mycronic is enhancing its semiconductor Pick and Place machines with AI, but effectiveness relies on access to large, sensitive datasets owned by their customers.
Federated Multi-task Learning: Jun 14, 2024; Federated Multi-task Learning combines federated and multi-task learning, where clients train local models and share only necessary parameters to maintain privacy across heterogeneous systems.
Enhancing data security with trusted execution environments: May 21, 2024; Trusted Execution Environments provide secure hardware-based protection for code and data. Benchmarks show Intel SGX excels with smaller applications, while AMD SEV handles larger workloads better.
Email Spam Detection with FEDn and Hugging Face: May 17, 2024; Using the Hugging Face Transformers library in FEDn to fine-tune a BERT-tiny model for spam detection on the Enron email dataset, achieving ~99% accuracy with federated privacy.
Federated Self-supervised Learning and Autonomous Driving: May 13, 2024; Federated self-supervised learning trains AI on autonomous vehicles using sensor data without centralizing or labeling it, handling privacy, compliance, and data volume challenges.
Leveraging JWT Authentication for Secure Client and Admin API Access in FEDn Studio: May 8, 2024; Using JWT authentication in FEDn Studio to secure API access for clients and admins, ensuring only authenticated users can access and manage federated projects.
Simplifying Federated Project Management with ArgoCD in FEDn Studio: May 8, 2024; How ArgoCD enhances FEDn Studio for managing federated projects on Kubernetes, using a custom helm chart for automation, improving deployment efficiency and scalability.
The impact of the backdoor attack: May 7, 2024; Backdoor attacks in federated learning, with experiments on the MNIST dataset demonstrating the challenge of detecting attacks when a minority of clients insert hidden triggers into the data.
Federated Learning: Self-managed On-premise or SaaS?: May 2, 2024; Scaleout's FEDn offers SaaS and self-managed deployment options for federated learning, providing flexibility, security, and scalability to meet evolving user needs.
Scaleout and Flower partner on federated learning solutions: Apr 22, 2024; A strategic collaboration with Flower enables developers to run Flower projects on FEDn, providing access to enterprise-grade security, scalability, and monitoring capabilities.
Input Privacy: Adversarial attacks and their impact on federated model training: Mar 25, 2024; Examination of label-flipping attacks on federated machine learning, showing limited impact on the global model's accuracy due to federated averaging limiting malicious clients' influence.

Deploying AI at the Edge?
Discuss your use case with us.

Talk to us Download the platform overview