Tim Slusser, Director of Smart Mobility Initiatives
The tragedies of Lion Air flight 610 and Ethiopian Airlines flight 302 caused both mourning and concern on a global scale with passengers and airlines alike seeking answers. The potential cause of these crashes has been extensively reported in the media as a failure of the Maneuvering Characteristics Augmentation System (MCAS). Since automation plays such a large role in the world of smart mobility technology, I’d like to share my thoughts on how this system and its failure are related specifically to the development of autonomous vehicles (AVs).
Similar to the technology being developed for both fleet and personal vehicles, the MCAS system was designed as an automated “driving” system for the 737 Max. The software was developed by Boeing engineers to automatically detect dangerous climbing angles, and via inputs from the sensor system, execute flight maneuvers to counteract them. In the case of the Lion Air and Ethiopian Airlines flights, it is believed that the sensors went bad and the software received inaccurate data.
The direct link to AV development on the ground is that these vehicles utilize the same types of systems – sensors and software. The software is written by human beings, the sensors are a developing technology, and there are infinite driving scenarios that exist. The magnitude of this engineering challenge cannot be underestimated. For some perspective, approximately 30-40 years ago we were writing software that displayed “Hello World” on black and white (or maybe green) computer screens. Now we are trying to create an artificial intelligence that can understand environmental inputs that might not even exist today (e-scooters hit the road just a few years ago) and execute high-speed driving maneuvers based on those inputs.
Fortunately, this challenge hasn’t escaped the vast majority of AV developers. While deployments have not yet achieved the lofty promises of years past, that’s actually a good thing. Technology developers are taking a slow and methodical approach with a high regard for safety. A quick review of AV deployments to-date will show you that active vehicles are limited to low speeds and geofenced areas. This approach provides the opportunity to engage with consumers, collect massive amounts of data, and improve AV system performance, all while limiting safety risks.
In addition to the many technology improvements needed to make “AVs-for-all” a reality, there is a great deal of work to be done to engage and prepare cities and communities. Community leaders and stakeholders need confidence that public safety won’t be compromised, to understand how the technology will provide new benefits to residents and visitors, and prepare for costs associated with deployment. Efforts that require active local participation and explicit approval, such as installing ground-based sensors, can also help to accelerate AV development.
This type of local engagement is quietly happening in cities and communities across the world. AV developers such as Michigan-based May Mobility have already had the thorough conversations with cities like Detroit and Grand Rapids ahead of activating their fleets on city streets. And while the fleet deployments are always covered in the media, the importance of city/community participation is often overlooked or omitted. Since these types of fleet deployments are targeted at addressing local needs, and those local needs will be unique to every location across the globe, these conversations need to continue.
To learn more about how NextEnergy is helping our partners achieve the goals of an autonomous vehicle future, and how we can help accelerate your smarter, cleaner, more accessible mobility solutions, feel free to reach out to me directly.