Autonomous Business Process (ABP) management is an approach to business process governance that seeks to optimize process workflow without human assistance. ABP intelligent software agents, also known as bots, are designed to independently discover and map business processes and use machine learning (ML) algorithms to improve process flow.
Although it is still an emerging technology, mature applications are starting to emerge, such as:
In aviation, ABP software can autonomously determine the role of people, technology, and data in a particular business process and independently figure out how to optimize it. And some diffuse applications are appearing in particular areas such as airports and airlines management. Let’s hope that the progress made in other industries soon encourages aviation to explore new AI-powered initiatives.
With the mainstream adoption of assistants such as Siri, Alexa or “OK Google”, voice-based assistants have infiltrated our everyday life. Due to their overwhelming success, that trend is probably not going anywhere.
More dedicated recommendations algorithms are being implemented to an extent that these assistants will try to provide you with personalized assistance. For example, bots are starting to provide legal advice, business decisions, or even medical assistance.
In aviation, AI assistants are still cutting-edge technology. Some airlines have started to leverage flight booking applications with Alexa skills or Google apps, yet nothing very fancy is coming out of it. However, some projects are looking to introduce voice assistants in the cockpit. That said, regulations and certifications have barred much progression. Let’s see if 2022 is a good year for AI assistants in Aviation.
Huge advancements in the automobile industry are expected in the coming years. Autonomous cars are already being tested on streets, despite considerable debate on their safety. Seems that “ethical ML” is going to be the hot topic this year.
In the Aviation context, there have already been rumblings of introducing single-pilot operations for freight and short-range flights. Also, there are some problems that seek the use of AI for taking off and landing without a runway, seeking out obstacles (like vehicles, buildings and birds), or altering course to manage unpredictable situations (like wind gusts, engine failures and obstacles)
Even though Deep Learning has thrived in a variety of industries, one of the technology’s limits has always been its dependency on large amounts of data and computer power. One of the obvious limitations of today’s DL is its reliance on huge amounts of labeled data and computing power.
Self-Supervised learning is a promising new technique in deep learning in which, instead of training a system with labeled data, it is trained to self-label the data using raw forms of data. Any input component will be able to anticipate any other part of the input in a self-supervised system. It might, for example, forecast the future based on the past.
The aviation industry is notorious for its lack of labeled historical data. In this context, self-supervised learning algorithms can be the key to tackling problems such as Flight Data Monitoring (FDM) outliers detection or most probable flight plan forecasting.