Embedded World 2026

In today’s rapidly evolving world, organizations must adopt AI for their edge deployments to stay competitive — yet modernizing mission‑critical systems while adopting cloud‑native and edge AI technologies is complex and risky. Drawing on more than 40 years of enabling and powering mission‑critical systems in secure and safe highly regulated markets, Wind River is at the forefront of edge AI transformation, with a reliable, end‑to‑end foundation for edge AI use cases that preserves security, safety, and reliability throughout the lifecycle.

Maintaining ROS on Yocto: A Survival Guide for Embedded Open Source Integrators

Speaker: Rob Woolley, Senior Principal Technologist

Abstract: Maintaining ROS on open embedded/Yocto is a complex task that blends technical challenges with communication and leadership skills. This includes mixing robotics and embedded Linux with setting up infrastructure and encouraging participating. This talk offers a deep dive into the challenges of creating a project at the intersection of two other large open source communities. We’ll discuss integrating ROS into deeply embedded environments, including improving the workflow for those unfamiliar with embedded, and welcoming both junior and senior contributors. We’ll also share practical strategies for managing long-term maintenance, collaborating with upstream communities, and ensuring reproducibility across diverse hardware targets. Attendees will leave with actionable insights into how to sustain open source projects while maintaining their full-time career.

Real-Time Meets Cloud: Orchestrating RTOS and Linux with Kubernetes

Speaker: Andrei Kholodnyi, Principal Technologist

Abstract: Modern embedded systems often require deterministic real-time performance alongside scalable cloud-native applications. This talk explores deploying workloads across the real-time operating system (RTOS) VxWorks^® and Linux environments within a Kubernetes cluster, using ROS 2 as a concrete example of a distributed application. We’ll demonstrate how OCI-compliant containers can be built for both an RTOS and Linux, enabling mixed-platform orchestration with real-time tasks coexisting with general-purpose cloud workloads. A recorded demo will showcase ROS 2 nodes running seamlessly across RTOS and Linux nodes, illustrating practical strategies for bridging embedded determinism with cloud scalability and unified management

Industrializing CHERI on RISC-V: VxWorks and Hypervisor Support for Safety-Critical Edge Systems

Speaker: Dmitriy Yeliseyev, Software Architect

Abstract: Memory safety violations remain the leading cause of software vulnerabilities in embedded and mission-critical systems. Capability Hardware Enhanced RISC Instructions (CHERI) provide a hardware-enforced model for fine-grained memory protection, significantly reducing the risk of exploitation while allowing existing software to be adapted rather than rewritten. This makes CHERI particularly attractive for operating systems such as VxWorks^®, which demand that security and reliability be achieved without sacrificing real-time determinism.

Building on earlier work with CHERI on Arm’s Morello board, we present our extension of these concepts to RISC-V architecture. RISC-V is rapidly emerging as a leading open and scalable processor ecosystem, with growing industrial traction across aerospace, defense, automotive, and telecom markets. The Codasip X730 SoC provides a concrete and forward-looking platform, with LLVM17 toolchain support and the potential to serve as a reference point for industrial deployment.

Our work establishes a CHERI-enabled VxWorks software stack on RISC-V by upgrading the LLVM toolchain, developing a new board support package for the X730 and enabling CHERI awareness across the operating system core, kernel, and user-space layers. Validation is carried out in both emulated and hardware environments, with emphasis on certification-friendly engineering practices.

We argue that CHERI can be industrialized beyond research prototypes, bridging academic exploration to commercially viable solutions. By proving feasibility on both Arm and RISC-V, we aim to accelerate adoption of CHERI as a foundation for the next generation of secure, connected, and safety-critical systems.

Web Assembly in Safety-Critical Embedded Systems: A Runtime for the Heterogeneous Future

Speaker: Dan Milea, Sr. Manager, Engineering – Sys

Abstract: Emerging heterogeneous embedded platforms integrate CPUs, GPUs, FPGAs, and NPUs to address the stringent demands of real-time AI, control, and safety-critical applications. Yet current software stacks may lack a unifying abstraction that ensures portability, predictable execution, and safety across such diverse compute resources. WebAssembly (Wasm) is gaining traction as a lightweight runtime technology offering platform-independent binaries, memory safety, and a sandboxed execution model that inherently reduces attack surfaces.

Recent advances such as WASI (WebAssembly System Interface) and component models enable fine-grained modularity, dynamic linking, and deterministic execution, making Wasm suitable for embedded domains with real-time and mixed-criticality requirements. Its isolation properties simplify safety argumentation under standards such as ISO 26262 and DO-178C while supporting incremental certification through modular deployment. Furthermore, Wasm can integrate with Kubernetes-based orchestration layers, enabling consistent workload scheduling, lifecycle management, and adaptive scaling across constrained devices and cloud-connected systems.

This convergence of Wasm runtimes, safety frameworks, and orchestration tooling positions WebAssembly as a key enabler for real-time, secure, and portable execution on heterogeneous embedded platforms. The presentation will detail runtime architectures, scheduling strategies, and safety benefits and will demonstrate how Wasm can underpin the next generation of orchestrated embedded deployments.

From RTOS to Linux: Navigating OS Options for the Intelligent Edge

Speaker: Michel Chabroux, Vice President Product Management, Wind River

Abstract: In the evolving OS landscape, decisions go beyond tech specs. While Linux dominates openness and versatility, embedded and edge use cases demand more nuanced choices. Factors such as resource efficiency, lifecycle costs, compliance, and time-to-market are key. AI workloads add complexity, requiring optimal compute, real-time capabilities, TSN, and AI framework support. Attendees will learn to evaluate OS options — from RTOS to Linux — based on practical needs, balancing performance, security, and market demands over ideology.

From Silos to Synergy: Enabling Manufacturing Excellence Through IT/OT Convergence

Speaker: Sandeep Modhvadia, Chief Product Officer

Abstract: As manufacturing advances into the Industry 4.0 era, the lines between information technology (IT) and operational technology (OT) are rapidly dissolving. By 2025, more than 75% of leading manufacturers will have embraced IT/OT convergence to unlock double-digit efficiency gains, minimize downtime, and accelerate innovation. Yet many still face the barriers of legacy systems, siloed operations, and escalating security and management complexity — challenges that can stall transformation and erode potential ROI.

This session explores how leading manufacturers are bridging IT and OT using modern infrastructure platforms purpose built for distributed, mission-critical environments. Attendees will discover proven architectural patterns, deployment strategies, and organizational approaches that enable seamless convergence — from the intelligent edge to centralized cloud orchestration — without compromising uptime, deterministic performance, or security. Beyond technical innovation, the discussion centers on measurable business impact: lowering total cost of ownership, optimizing asset utilization, and driving new revenue streams through smarter, data-driven operations.

Wind River^® Cloud Platform serves as a unified infrastructure capable of hosting and supporting both IT and OT workloads, simplifying operations while delivering scalability, consistency, and financial efficiency. Wind River Conductor manages the full lifecycle of this unified platform, orchestrating workloads across environments to ensure continuous performance and compliance while reducing manual effort and operational expenditure. Conductor’s automated blueprints accelerate provisioning across on-premises and hybrid clouds of any size, enabling cost savings through automation and faster time-to-value. Meanwhile, critical OT functions such as Wind River Studio Analytics and other Cloud Platform services can now operate directly at the edge, bringing intelligence closer to production and creating opportunities for revenue-generating innovation. Extending these capabilities even further are VxWorks^® and Wind River Linux, which connect industrial devices and systems directly to Cloud Platform — delivering a unified, cost-efficient, and performance-optimized foundation for next-generation industrial operations.

AI Agent-Driven Simulation Using Natural Language Interaction with vECUs

Speaker: James Hui, Product Line Manager

Abstract: AI agents integrated with simulated virtual targets enable a new way to model and analyze a full embedded system, such as virtual ECU that consists of multiple microcontrollers connected via a CAN network, without physical hardware. Users can configure, monitor, and debug these complex environments using natural language, eliminating the need for deep simulation expertise. AI agents can automate such tasks as setting up CAN nodes, injecting bus faults, tracing message flows, and generating diagnostic scripts for anomaly detection. This approach accelerates experimentation, reduces cost, and supports rapid iteration for safety-critical and IoT applications. By fostering cross-discipline collaboration and creating shared, reusable knowledge, teams can validate designs earlier, improve reliability, and innovate faster.

Nuremberg Messe

Hall 4 / Booth 4-112 March 10–12, 2026