Apex.AI, the company leveraging open-source code to develop breakthrough safety-certified software for the automotive industry, announced today that it has partnered with Toyota’s Woven Planet Group to help the company develop and deploy a production-ready autonomy stack for Arene, Toyota’s vehicle development platform to enable modern software development tools and best practices in the automotive industry. Toyota’s Woven Planet is using Apex.OS, an SDK (software development kit) for automotive and other safety-critical applications, including autonomous driving software.
Adding to Apex.AI’s momentum is a business partnership with Tier IV aimed at combining Tier IV’s Autoware-defined reference design and integration capabilities with Apex.AI’s software stack for safety-critical autonomous systems.
Both announcements represent significant moves away from today’s hardware-defined vehicles that leverage existing advanced driver assistance systems (ADAS) that have dozens of components that are not integrated with each other, say APEX.AI execs. They believe that the industry’s transition to autonomous, connected, electric, and shared vehicles requires software-defined vehicles with a much higher level of software integration, but that is creating hurdles for traditional automotive players.
A team of autonomy experts
Apex.AI was cofounded in 2017 by autonomous industry experts Jan Becker, now CEO and President, and Dejan Pangercic, now CTO, to help accelerate the development of software-defined vehicles. The company is backed by financial investors that believe in that mission, like Lightspeed Venture Partners and Canaan Partners as well as strategic investors Toyota AI Ventures; Volvo Group; Jaguar Land Rover’s venture capital arm, InMotion Ventures; Airbus Ventures; and Hella Ventures. In addition to Toyota and Tier IV, customers include GM Cruise’s Voyage.
Becker has been working on autonomous driving for 23 years, is an original author of the SAE autonomy levels, and lectures at Stanford University on autonomous vehicles and driver assistance. Previously, he served as the head of autonomous driving at Bosch North America and was a member of the Stanford Racing Team for the 2007 DARPA Urban Challenge.
Pangercic has been working on robotics and software frameworks since 2008. During his Ph.D. work at TU Munich, he developed ROS (Robot Operating System)-based software, and after joining Bosch in Palo Alto, he worked closely with Willow Garage on the core ROS codebase.
They lead a team that includes Sanjay Krishnan, VP of Product; Serkan Arslan, VP of Business Development and Sales; Zion Maffeo, VP of Legal and Business Affairs; Mehul Sagar, VP of Functional Safety; and Michael Pöhnl, Senior Staff Engineer, at Apex.AI GmbH.
Scalable and affordable vehicle SDK
According to a recent McKinsey report, “automakers have traditionally viewed software as secondary to hardware. They must now revisit this perspective, as well as their development approaches since software is now a prime value driver in the product development portfolio.”
With the increasing complexity of today’s software applications, current solutions aren’t scaling, and piecing together antiquated systems comes at a much higher cost of integration time, money, and performance, says Apex.AI. The company is providing OEMs and Tier 1 suppliers with a single Apex.OS SDK that can address all vehicle’s software needs so that every component can work harmoniously, providing a powerful, scalable, and affordable solution.
Much like a smartphone SDK, Apex.OS enables non-expert developers to create reliable and complex applications by leveraging APIs (application programming interfaces) to take the complexity out of writing automotive software with simpler interfaces. It is based on the open-source ROS (Robot Operating System) and adds real-time processing, functional safety, performance, and support for commercial and safety-critical products. Most automotive companies are using ROS for R&D projects, but the software isn’t sophisticated enough for production vehicles.
Apex.OS bridges the gap between R&D and production-quality solutions by overcoming the real-time and reliability challenges. Switching to the OS means an easy transition for engineers, without any of the issues of legacy systems. Its value proposition is that the company handles complex, tedious, and safety-critical programming so developers can focus on the real-world problems they are trying to solve, like autonomous driving.
ISO certification gives Apex.OS a boost
Reinforcing the importance and legitimacy of the Apex.OS is that it is now TÜV NORD certified for functional safety to the highest level of ISO 26262, making the technology verified production-vehicle-ready. The certification is ASIL D safety level, which is an automotive risk classification that looks at the functional safety requirements for all the electrical, electronics, and software systems in a vehicle. It represents the highest level of risk management according to the most stringent safety requirements and targets less than 10 failures in one billion hours of operation for hardware components.
“It was a longstanding belief that open-source code was not certifiable; however, our team achieved certification in record time so that it conforms to the highest requirements of the applicable automotive functional safety standard,” said Becker. “If a software crash happens on your laptop it’s inconvenient, but if the software crashes in any safety-critical function of a vehicle it can be catastrophic. This is why we set out to write reliable software that protects against system crashes or operation failures. The certification proves we accomplished our goal as our software targets failure rates so low that they cannot be expressed statistically.”
Enabling Toyota’s innovative vehicle-development platform
Toyota’s Woven Plant recently detailed its partnership with Apex.AI via a customer success story.
“Arene is our vision for the most programmable vehicle on the planet,” said James Kuffner, CEO of the Woven Planet Group; Board of Directors, Toyota Motor Corp. “Apex.OS embodies our vision of open APIs for developers to weave mobility and autonomy into our daily lives. It enables agile development and deployment for engineers, shortening the time from concept to deployment, and making rapid innovations in autonomous driving possible at the scale of Toyota.”
Arene is a vehicle development platform with state-of-the-art tools, vehicle application programming interface (APIs), and safety building blocks that allow rapid iteration to shorten the time from concept to deployment. Apex.OS is integrating into Arene’s software stack to shorten the R&D of autonomous software development, enabling Toyota to create production-grade prototypes at a record pace.
“Toyota is a pioneer in autonomous mobility, and Apex.OS is accelerating innovation within the industry by providing a robust and reliable SDK that enables rapid, scalable, and affordable autonomous system development,” said Becker.
Tier IV and Apex.AI
The business partnership with Tier IV, a tech startup in Japan known as the original creator of Autoware, is expected to result in the world’s first open-source software for autonomous driving. Examples of its application include a driverless robo-taxi program launched in Tokyo. The partnership is aimed at combining Tier IV’s Autoware-defined reference design and integration capabilities with Apex.AI’s software stack for safety-critical autonomous systems.
The partnership of Tier IV and Apex.AI is said to mark a significant milestone in mobility technology. The companies are creating a path to production by leveraging open-source components while capturing a wider range of safety requirements from the market through various offerings jointly developed by the partnership.
“We are leaders in the ecosystem of companies building autonomous-driving cars with ROS and Autoware,” said Shinpei Kato, Founder and CTO, Tier IV. “With this partnership, our OEM and Tier 1 customer projects in Japan will benefit from global expertise in state-of-the-art software development, with integration and deployment support from Tier IV.”
In 2018, Apex.AI teamed with Tier IV on the Autoware.Auto to successfully demonstrate autonomous driving on the streets of Palo Alto. The solution combines state-of-the-art architectures, algorithms, software design, implementation, and integration techniques, with sensing, localization, perception, planning, control, and mapping components based on ROS 2. By supporting the commonly used sensing modalities including camera, radar, LiDAR, IMU, and GNSS in a sensor-agnostic and ECU-agnostic way, advanced sensor fusion approaches can be implemented for higher accuracy, robustness, and wide applicability.
