How do you ensure an airliner can land autonomously at any airport at any time using computer vision instead of pilots’ eyes? The answer involves collecting and processing millions of real-world images, in dierent locations and conditions, to train a machine-learning algorithm that is robust and reliable. That is why Airbus’ Silicon Valley outpost Acubed is performing hundreds of fl ights of a Beechcraft Baron 58 test aircraft under the Wayfi nder project. It is developing the core technology for certifi able autonomous fl ight systems that could be used across Airbus’ next generation of airliners .
展开▼
