Aviation pilots' challenge: weather awareness and decision-making

The situation

In aviation, weather is one of the leading causes of flight delays and safety incidents. Specifically, convective weather (e.g., thunderstorms) is very challenging, because of its associated hazards (e.g. wind shears, turbulences, hail). Several discussions with expert pilots also suggest that its vertical characteristic is not very well understood because of the lack of weather-related knowledge and skills that pilots acquire during their training process. In addition, current operational ground-air technologies and services limit pilots’ weather-related situation awareness and decision-making capabilities, and can lead to deadly accidents.

My role, team and responsibilities

As the main researcher and designer, I led the full research lifecycle over three and a half years. Under the guidance of my Ph.D. supervisor, I collaborated with expert pilots, meteorologists, interaction designers, cognitive psychologists and interns. I was responsible for uncovering how pilots manage weather during a flight and provide ways to improve their situation awareness and decision-making for safer, more efficient and comfortable flights.

A pragmatic, non-linear research process

I was very new to the world of aviation weather. I had to be pragmatic and open-minded. I first took weather classes to better understand the foundations. I started talking with expert pilots to better understand weather in the context of aviation. As my knowledge got better, I refined my questions to the experts, prototyped ideas and concepts to gather their feedback, and eventually made good progress with the work! My research was deeply grounded in qualitative research and iterative prototyping to better understand how pilots perceive, understand and make weather-related decisions throughout a flight. I explain the high-level story down below.

Grounded theory research

I conducted in-depth, semi-structured interviews with 15 expert airline pilots, to explore how they build weather-related situation awareness from pre-flight briefings to in-flight decisions.

I identified recurring patterns, mental models, associated challenges and needs. The data mainly pointed out that existing onboard weather tools often rely on 2D displays and limited interaction capabilities, offering little support for pilots to develop an accurate, real-time mental model of weather conditions.

At this point, the main question I had was: how can interactive 2D/3D weather information improve airline pilots’ situation awareness and in-flight decision-making? The goal was to create a tool that pilots would trust, use, and understand intuitively under pressure.

Participatory design

Based on insights I uncovered from my interviews, I started to work on how to visualize weather. Instead of designing in isolation, I brought the evolving design to the users early in the design phase. I organized a focus group with two experienced airline pilots, alongside our interdisciplinary design team. The goal was to validate the best weather representation approach.

The discussion was revealing. Pilots spoke about the cognitive load they face in adverse weather, and how visual clutter can actually worsen their decision-making. They guided us toward simpler yet information-rich visualizations: 3D storm cells that were color-coded, semi-transparent, and responsive to vertical slicing for altitude-specific views.

This moment in the design process ensured that the representations moving forward weren’t just engineered; they were understood by expert pilots. That focus group helped crystalize key design choices and ultimately improved its clarity and usability in simulation tests that followed.

Prototyping and testing

Over three years, I led the development of three major prototypes of what I called the Onboard Weather Situation Awareness System (OWSAS), progressively refining interaction models, visual representations, and system architecture. Evaluations were conducted with expert pilots using high-fidelity Boeing 737-800 cockpit simulator. Main human factors evaluation metrics covered workload (NASA-TLX), usability (SUS), cognitive compatibility, and decision performance along with additional pilot feedback.