• Fadi Labib

Apex.OS 1.3 Release

Updated: Aug 4

Apex.OS 1.3 is now available! We walked (actually, drove) many miles in our customer’s shoes and spent the past few months making Apex.OS an extremely developer-friendly framework for creating modern, automotive grade software. Recognizing our work, the German Design Council awarded Apex.OS with two awards in June: Best of Best (Mobility & Innovation Category), and Innovation of the Year (Special Category).

Node graph of an advanced automotive software system

Our customers are developing real world applications for Advanced Driver Assistance Systems (ADAS) or Autonomous Driving (AD) with multiple sensors and software components. These applications are architected as complex graphs of nodes with a combination of data-driven and time-driven workflows that need to run in a deterministic and real-time manner. To achieve that, Apex.OS introduces a deterministic executor with this release to abstract the complex scheduling policies of various Linux / RTOS flavors and simplifies the development of a safety certifiable software stack.

Deterministic execution of C++ applications has been an active research area, with the robotics and automotive industries leading the development of such capabilities. We are excited to bring this feature to our customers, providing a viable route to safety-critical deployment on any hardware and OS of their choice.

What if these software components are running on different processors that need to share Gigabytes of data with microsecond latencies? Apex.Middleware has you covered. Apex.Middleware, introduced in this release, can be used with any development framework, including Apex.OS and AUTOSAR Adaptive. Built as a highly optimized integration of Eclipse Cyclone DDS ™ and Eclipse iceoryx ™, Apex.Middleware was designed to be faster than an express train, and leap over tall buildings. With zero copy communication, transport agnostic Quality of Service (QoS) settings, DDS security, dynamic discovery, and wire compatibility with RTPS, SOME/IP and MQTT, Apex.Middleware gives your application superpowers!

Another hugely important need that our customers have is the ability to deterministically replay the data from their systems under test. This would allow testing and validation to happen, which is necessary for deploying automotive grade features on public roads. You guessed it - this release of Apex.OS makes it possible to deterministically replay! Using this capability of Apex.OS, it is now possible to replay a simulated or live recorded scenario and be assured that you will observe the same behavior every time. Validating a software stack to reach ISO 26262 automotive standards is now a great deal easier.

In making all this happen, we also contributed features and performance improvements to the open-source ROS 2 (rosbag2) project. The Apex.Middleware effort included significant contributions to the Eclipse Cyclone DDS and Eclipse iceoryx projects. We are pleased to be able to contribute to these open-source projects, and thank the community for helping make it possible for enthusiasts and large companies to design and deploy advanced applications without re-inventing the core building blocks.

Finally, we added a number of articles for Apex.OS users to be even more productive. Included in this release are a step-by-step guide to porting code from ROS or ROS 2 to Apex.OS, an article about defining and using messages to interface with sensors, and a detailed description of how to use the zero copy feature to achieve ultra low latency with large data such as camera and lidar images.

If you are interested in using Apex.AI products for your projects, please...

We are also growing our team worldwide. Please see open positions on our careers page.


Recent Posts

See All