Chris
By Alan Muscat
Aviation has always been at the forefront of innovation and, as technology advances, the role of the pilot is increasingly becoming supervisory in nature, raising new challenges for maintaining safety and operational awareness. The Artificial Intelligence-based Additional Pilot (ArtiAP) project – spearheaded by researchers at the Institute of Aerospace Technologies at the University of Malta and aviation subject matter experts from QuAero, the industrial partner in the project – aims to address these challenges by exploring how Artificial Intelligence (AI) and Machine Learning (ML) could support pilots in complex and high-workload scenarios in the future.
As the industry evolves, the traditional reliance on human decision-making and situational awareness in the cockpit is being augmented by advanced automation technologies. ArtiAP aims to complement crew capabilities by providing tailored assistance that adapts to their needs, including offering additional support to less experienced crews or those new to a specific aircraft type. Furthermore, ArtiAP can provide critical assistance in scenarios where one crew member is incapacitated, ensuring seamless operation and effective decision-making.
At its core, ArtiAP envisions an intelligent, human-centric system that enhances situational awareness and decision-making without removing pilots from the control loop. The AI technology under development is designed to assist pilots during routine and abnormal operations e.g. by providing support during en-route diversions and during the execution of emergency descent procedures in mountainous terrain. Through visual and aural alerts, as well as multi-modal interaction via voice commands and touchscreen gestures, ArtiAP strives to integrate seamlessly into the cockpit environment. Importantly, the system is intended to work in collaboration with pilots rather than as a replacement, preserving their central role in decision-making.
Key to the project is the use of Generative Artificial Intelligence (GenAI) with Large Language Models (LLMs) that provide a natural and intuitive interface between the flight crew and the on-board automation. With this technology, pilots can interact with cockpit automation via voice or text using natural language commands, as if they were talking to a human being. For instance, a pilot could ask ‘What is the weather like at our destination?’ or ‘What is the probability that we will have to hold at the destination?’
The project is adopting a phased approach. It began by identifying key use cases where AI-driven assistance could have the greatest impact, such as mitigating high workload conditions or improving response time during emergencies. Using advanced ML techniques, ArtiAP is developing, testing, and refining ML models tailored to these scenarios. In addition, the project is designing a user-friendly Human-Machine Interface (HMI) to enable effective communication between pilots and the AI.
Building on prior research efforts like TOUCH-FLIGHT2/ePM and SMARTAP, ArtiAP also aligns closely with European goals for sustainable and AI-integrated aviation, as outlined in documents such as EASA’s “AI Roadmap” and ACARE’s “Fly the Green Deal”. Ultimately, the technology will be evaluated in a flight simulation environment, incorporating input from airline pilots to assess its usability and effectiveness.
ArtiAP (R&I-2021-014-T) is a €295,000 project financed by the Malta Council for Science & Technology, for and on behalf of the Foundation for Science and Technology, through the FUSION: R&I Technology Development Programme.
Marvic
