McKinsey Webinar: Exploring Impact of Machine Vision on Aviation
Advancements in artificial intelligence (AI) and machine learning are enabling autonomous technology to become increasingly useful in both traditional aviation and emerging aviation fields. McKinsey & Company, an industry-leading consulting firm, hosted a recent panel on how machine vision is changing the aviation industry. The panel was led by Robin Reidel, a McKinsey & Company partner and distinguished voice in aviation. I was honored to represent Acubed alongside panelists from AeroVect, Assaia, and Skydio to discuss how each of our companies is leveraging machine learning in a unique way, from baggage handling to autonomous flight.
Machine Vision in Aviation
Machine vision integrates computer vision – the use of images and 3D signals to interpret data – with a host machine or system to complete real-world tasks. While machine vision is not a new concept, per se, increased computational power and other technological advancements are making it increasingly reliable and beneficial.
In aviation, machine vision technologies are being leveraged to support operations management at airports, enable self-driving ground support equipment (GSE) and facilitate autonomous flight both in uncrewed aerial systems (UAS) and traditional aircraft, among other applications.
Assaia, for example, offers computer vision software that monitors what is happening on an airport’s apron during aircraft turnaround. The company’s software makes apron operations transparent, allowing airports and airlines to better manage them and implement changes to improve safety, efficiency, and sustainability. According to Assaia’s co-founder and CEO, Max Diez, because airports function as an ecosystem – where delays can snowball – these changes can improve overall efficiency.
Machine vision-enabled GSE vehicles, like those developed by AeroVect, are increasingly part of that ecosystem too. AeroVect takes existing vehicles from customers – vehicles like baggage and cargo tugs – and outfits them for autonomous use in airports, improving safety and limiting the risk of human error and delay. Raymond Wang, AeroVect’s co-founder and CEO, says these improvements are not five or 10 years away – they are happening now. (The company has already mapped some of the world’s largest airports and deployed autonomous vehicles in several U.S. locations.)
Why Now?
This is an exciting time for AI and machine learning due in part to a tremendous uptick in machine vision-enabled products. During the panel, we discussed how this uptick is the result of increased computational power, the availability of data, and the ability to process that data using advanced algorithms. Panelists agreed that this explosion of deep learning has been fundamental to machine vision’s increasing functionality in a variety of fields.
Skydio co-founder and CEO, Adam Bry, says that the technology and products are now able to effectively scale and serve a range of industries. Skydio, a leading drone manufacturer making autonomous drones, has so far applied its technology to a wide range of applications from cinematography to inspections to defense.
Wayfinder & Autonomous Flight
At Acubed, we are focusing our machine learning efforts and leveraging them for commercial aircraft autonomy. We have now achieved breakthroughs in taxi, takeoff, and landing for large commercial aircraft, increasing the safety and efficiency of aircraft operations.
Our development of this machine vision technology is part of the company’s portfolio. Using a Beechcraft Baron 58 modified with sensors and cameras, We are now gathering real-world flight data by visiting airports throughout the United States. Leveraging this technology, we achieved world firsts in 2020 by autonomously taxiing, taking off, and landing a commercial aircraft in Airbus’ Autonomous Taxi, Take-Off and Landing (ATTOL) project. Building on ATTOL, the DragonFly project was launched in 2022 to perform autonomous emergency operations in the event of crew incapacitation.
Future of Flight
For us, these milestones are a first step to help enable certification for the next generation of aircraft and they are not simply technology for technology’s sake.
There is a huge demand.
Experts expect flight passengers to increase to more than eight billion by 2037. To respond to this growing demand and ensure a safe rollout of new autonomous technologies, Acubed and Airbus work with regulatory agencies like The European Union Aviation Safety Agency (EASA). Technological advancements like automation have contributed to increased safety in aviation and, with increasing flight demand, AI and autonomous systems are a way to make flight safer for pilots as well as passengers. (Wayfinder technology, for example, can be used if pilots find themselves in especially difficult situations.)
As to machine vision’s impact on the rest of the aviation industry? We all agreed: AI technology is accelerating and will be even more impactful that people may realize.