VxWorks：创新之旅

AI/MLPandas for datamanipulation; TensorFlowLite for ML modelingRustProgramminglanguage supportWind RiverStudio DeveloperCloud-enabled toolingFirst to marketI/MLANumPy for datamanipulationSecure DevelopmentLifecycle (SDL)NIST SP800-218 forproduct developmentPythonProgramminglanguage supportExtensive BSP support & Safety Certification Leadership TSN on Virtual MachineZero performancedegradationDeployment and updatesKubernetes support GNAT Profor RustAdaCorepartnershipOCI-compliantContainersFirst to marketContainerOrchestrationFirst to marketDeveloperCertificationEnhancingdeveloperknowledge &efficiency FUELING HIGH PERFORMANCEEvery embedded systems project needs to deliver impressive results on time, within budget, and with innovativefeatures that help its users. VxWorks contributes to that effort.Bringing Artificial Intelligence and Machine Learning to the EdgeAI and machine learning (ML) workloads once were exclusively in the domain of the cloud. Today, AI and MLcomputations are made at the network’s edge, which reduces the data traffic sent to the cloud for processing.Milestone:VxWorks enables the rapid implementation of AI/ML via library extensions. TheseincludedNumPyin 2020 andPandasin 2021 to improve data manipulation. Also in 2021, weintroducedTensorFlow Litefor easier implementation of ML models.These enhancements equip development teams to capitalize on dynamic edge-to-cloud processing. For instance,using these features can help deployment of image recognition in aerospace and defense applications. It canprocess sensor data within a vehicle for predictive maintenance. And it can speed up medical diagnostics.Supporting Time-Sensitive NetworkingModular design is a popular approach to cost-effective systems development. System modules or componentsconnect via Ethernet, with the goal of interoperability across vendors’ equipment.However, traditional Ethernet connectivity cannot guarantee timely delivery of high-priority system data. Toaddress this issue, the IEEE industry standards organization (IEEE 802.1) is augmenting traditional Ethernetdefinitions to encompass Time-Sensitive Networking (TSN). Among other enhancements, the TSN WorkingGroup addresses time synchronization, scheduling, and traffic shaping.Milestone:VxWorks introduced TSN support in 2016. Since then, we have added support forArm®and Intel®hardware architectures. This lets embedded systems engineers minimizenetwork latency and jitter, providing bounded end-to-end delay and guaranteeing data deliverytime over Ethernet.Open Platform Communication Unified Architecture (OPC UA) is a cross-platform, open source, IEC62541standard for data exchange from sensors to cloud applications developed by the OPC Foundation. It provides amodern, secure, and reliable way to facilitate HMI/SCADA and general-purpose industrial automation activities, aswell as integration with business intelligence, analytics, and cloud platforms for Industry 4.0. Wind River is anactive contributor to the open62541 community.Milestone:Theopen62541(OPCUA stack) over TSN was added to VxWorks in 2020.High-performance processors increasingly carry on-board TSN capabilities. Each VxWorks release enhances TSNprocessor support to keep up with evolving TSN standards and to improve high-performance system designflexibility. Systems consolidation projects often place requirements on these processors to run multiple systemfunctions on multiple OSes, within virtual machines via a hypervisor.Milestone:March 2024 saw formal validation testing of VxWorks supporting TSN features as aguest operating system on Wind River HelixTMVirtualization Platform, our safety-certifiedhypervisor. This proved that there was no degradation of performance when VxWorks is used forTSN applications in a virtual machine (VM). In support of TSN, we participate inPlugFest workshops, industry forums, and collaborative activities with embed-ded systems vendors across industries.Learn more:Following close collaboration with aerospace and defense and automotivesystems vendors, we published a white paper, “Hitting the Moving Target with TSN,”which describes how development teams can move forward with TSN implementation intheir designs.Modular design has become commonplace in aerospace and defense systems, where it is crucial to source equip-ment from different vendors and to upgrade or replace modules efficiently. Automotive vehicle systems architec-tures are also adopting TSN to support the adoption of on-board, high-priority, safety-critical sensor data.ENABLING MORE COST-EFFECTIVE FUNCTIONALSAFETY CERTIFICATIONFunctional safety certification is expensive, risk-heavy, time-consuming — and absolutely necessary. VxWorkscertification evidence enables development teams to achieve cost-effective certification of designs so engineerscan focus on the code they build, not the tools they use.This is a key advantage with today’s embedded systems, which commonly include multi-core processors. Multi-coreprocessors present sev