Platform agnosticism: The key to growth in the autonomous vehicle industry

While research into autonomous technology has been going on for decades, we continue to discover new and better ways to build the systems and platforms that get you from point A to B. To function properly and perform maneuvers safely, autonomous vehicles (AVs) use a combination of sensors, computers and algorithms. Yes, it’s a bit more complicated than it sounds, but many AV companies complicate efforts further by designing these components to operate as a fully custom transit solution.

This can cause problems because full customization can be very expensive and time-consuming, thereby limiting the viability of AVs as public transportation. For example, designing and testing new sensors requires significant investment in research and development, as well as a lot of resources to bring the sensors to market.

We make a concerted effort to optimize each dollar spent and that’s why our autonomous software approach does not require custom sensors or computers. Instead, we aim to leverage high-quality, off-the-shelf components and remain platform agnostic.

For the industry to grow, we believe AV engineering requires a platform-agnostic approach. This refers to software that’s compatible with a broad range of operating systems and system hardware. By designing the system to be agnostic to platform software and hardware, our AV engineers can readily employ emerging technologies for safer, more reliable rides.

How does platform-agnostic software solve challenges to autonomous vehicle deployment?

Platform-agnostic software can play a crucial role in solving challenges to autonomous vehicle deployment by providing a standardized and interoperable software solution that works across different vehicle platforms and operating systems.

It can also provide a standardized communication interface that ensures that the different components can communicate with each other seamlessly. Secure communication between different components of the system is essential to the safety and reliability of the overall system.

Platform-agnostic AV software additionally allows altered components to remain compatible with the system as a whole. Therefore, parts can be upgraded, replaced with off-the-shelf alternatives or swapped out to tackle different environments quickly and less expensively. As the majority of the system is unchanged, the AV won’t need extensive retesting to confirm it’s safe for deployment.

May Mobility’s approach to platform-agnostic hardware

At May Mobility, each vehicle features our autonomous driving kit (ADK), made up of our autonomous driving node (ADN) computer and a platform-agnostic combination of sensors. Today, we use three types of sensors that work together to provide a complete picture of the environment with the capability to add more as needed in the future:

• LiDAR is a light-based sensor that measures distances and provides a 3D representation of our surroundings.

• Radar is a radio-based sensor that detects objects and their movements providing accurate distances and velocity.

• A camera is a visual-based sensor that captures images of the environment, giving us the ability to classify and differentiate between objects.

Each of these sensors has its strengths and fused together they provide the necessary inputs to our Multi-Policy Decision Making (MPDM) software.

Our ADK supports our MPDM platform, which takes in perception data and simulates thousands of possibilities every second. Using high-quality, platform-agnostic technology allows our AVs to see their surroundings accurately and react quickly, as well as anticipate potential hazards ahead of time. As a result, our AVs can safely navigate complex environments like Ann Arbor, MI, where pedestrians and complicated intersections are ever-present.

Overcoming the problem of sensor range and resolution

One of the biggest challenges in the development of AV’s using off-the-shelf technology is the limitation of the range and resolution of available sensors. As the speed of a vehicle increases, the sensors need to be able to gather data farther ahead to predict hazards and react in time. But as we extend the range of what a sensor can see, resolution decreases.

Weather conditions have a similar effect. Heavy snow or rain obscure the visual field of sensors. Therefore, the ability to interchange or upgrade sensors as their capabilities change is critical to the future of AV viability.

By implementing our ADK with our MPDM system, we address these issues in two ways.

1. Maximizing available information with multiple sensor types

Sensor redundancy is key to optimizing perception. By using cameras, radar and LiDAR together, our AVs are able to get a full understanding of their surroundings. We also aim to keep up to date with the latest sensor technology. As that technology improves, it enables us to stay competitive in the market by enhancing our AVs’ ability to handle increased speeds and poor weather. Sensor redundancy also allows our MPDM system to continue predicting hazards more effectively. This means our AVs can be deployed in many different environments without the need to redesign large systems on the vehicle.

2. Easy integration into our autonomous driving kit

Since our ADK is platform-agnostic, we can integrate new sensor technology without spending time and money to rebuild our system from the ground up. That also means we can move it from one platform to another while minimizing engineering effort and maintaining manufacturability. As a result, May Mobility’s autonomous vehicle technology is:

• Designed to leverage new technologies

• Applicable across a wider range of vehicles

• Deployable in a variety of settings and climates

• Efficient to deploy

• Cost-effective to produce and maintain

Platform-agnostic technology in action

Our highly flexible technology allows us to alter the vehicle very quickly. Keeping our component integrations simple improves our ability to quickly adapt and integrate our sensor stack across different vehicle platforms, thereby serving the community much faster and at a lower cost. By partnering with OEMs and using vehicles manufactured for AV tech integration, there are minimal delays to ensure our vehicles are safe to ride. We experienced these benefits as we shifted our autonomous vehicle fleet from the Lexus to the Sienna Autono-MaaS, and will continue to experience the benefits as we introduce new platforms in the future.

Most importantly, as we efficiently equip our vehicles, we will be better prepared to serve the communities we deploy in. In the last six years, our technology has helped us achieve successful AV deployments in 11 cities in the U.S. and Japan, with more on the way.

Unlocking industry growth with platform-agnostic technology

May Mobility’s approach to AV technology—both hardware and software—sets us apart in the industry. Our platform-agnostic approach has resulted in AVs and AV deployments that can adapt, improve and scale quickly, effectively and at more manageable costs for businesses and communities.