Case Study: Addepto’s AI Solutions for The Aviation Industry

The collaboration between Addepto and a leading global aviation technology company exemplifies a successful long-term partnership aimed at enhancing airport operations through advanced AI solutions. This comprehensive case study highlights three significant projects that collectively transformed the airport experience for both passengers and staff.

AI for the Aviation Industry

Addepto’s engagement with the Client began with a shared vision to innovate air transport communications and information technology. Initially, Addepto collaborated exclusively with the R&D unit responsible for developing experimental solutions. During this phase, Addepto’s role was focused on the research and development stage, without participation in production implementation.

As the partnership evolved, some of Addepto’s innovative R&D solutions were transitioned into production environments and implemented by in-house engineering teams. This success marked a turning point in the collaboration. In subsequent projects, Addepto’s role expanded significantly, encompassing the entire product lifecycle from initial R&D through to production deployment. This new approach involved close collaboration with Client’s engineers at every stage, ensuring seamless integration and optimal performance of the developed solutions.

The R&D solutions created by Addepto have since become the foundation for numerous production-ready implementations, demonstrating the long-term impact and scalability of their innovative work. Notably, the Digital Twin project served as a springboard for implementing further AI optimizations, including a system for optimizing flight planning based on specific business KPIs.

Building on these successes, the partnership has now focused on leveraging a wide array of technologies to create an assistance bot that streamlines airport operations and improves customer service. This comprehensive approach combines classical machine learning systems, rule-based algorithms, and cutting-edge Generative AI to develop a bot that significantly enhances the passenger experience in terms of service delivery.

The collaboration utilizes all possible technologies, including traditional machine learning, computer vision, and Generative AI, showcasing the real business value of applying AI solutions to complex aviation challenges.

Project Summaries

Intelligent Tracking System and Digital Twin Development

Client sought to replace its legacy barcode-based tracking system with an advanced image recognition system, which was a pivotal step in modernizing airport operations. Addepto played a crucial role in this transformation by developing AI modules capable of analyzing images from cameras throughout the airport, effectively integrating Digital Twin technology into the operational framework.

Objective

• Implement intelligent tracking solutions for luggage and create a Digital Twin of airport operations.

Challenges

• Build an infrastructure capable of processing real-time data from multiple sources.
• Provide a comprehensive view of airport operations to facilitate better decision-making.

Approach

• Data aggregation: Collected data from various operational systems, sensors, and cameras.
• Real-time information processing: Developed a platform that delivers operational insights through a 3D user interface, enabling users to interact with real-time data effectively.

Outcome

The integration of Digital Twin technology within this project represents a significant leap forward in managing airport operations. By creating a dynamic virtual replica of physical assets and processes, the Client can monitor every aspect of its operations in real-time. This comprehensive view allows for improved tracking of luggage and passengers while enhancing overall operational efficiency.

The Digital Twin acts as an overview of airport operations, enabling predictive maintenance and proactive decision-making. For instance, by simulating various operational scenarios within the Digital Twin framework, Client can identify potential bottlenecks in passenger flow or delays in baggage handling before they occur. This capability not only saves labor hours by automating routine tasks but also ensures that resources are allocated efficiently where they are needed most.

Furthermore, the Digital Twin integrates seamlessly with existing systems to provide insights into energy consumption trends across different areas of the airport. This holistic perspective enables management to optimize energy usage while maintaining a safe and welcoming environment for passengers and staff alike.

The partnership has established a robust framework for ongoing innovation in the aviation sector. By continuously integrating AI technologies alongside Digital Twin capabilities, Client has set a new standard for customer service in airports. This collaboration demonstrates how long-term partnerships can lead to significant advancements in operational efficiency and passenger experience.

Read the full case study: Improving internal processes at the airports

LLM-Based Assistance Bot for Airport Operations

Objective

The primary goal was to enhance the airport experience by providing timely information to passengers and reducing manual workloads for staff. This dual focus aimed to create a more efficient and enjoyable environment for all airport users.

Challenges

• Improve Access to Information for Passengers: Passengers often face confusion and stress due to the lack of readily available information regarding airport procedures, terminal locations, and flight statuses.
• Minimize Manual Interventions by Airport Personnel: Airport staff frequently find themselves overwhelmed with routine inquiries, which detracts from their ability to focus on critical tasks such as security and emergency response.

Approach

Developed a dual-use case approach targeting both staff and passengers.

Outcome

The project resulted in a cloud-based solution utilizing GPT-4 and advanced facial recognition algorithms. Although initially envisioned as a humanoid robot, budget constraints led to a virtual assistant design. This innovative approach significantly improved operational efficiency and passenger satisfaction.

Read the full case study: LLM-based Assistance Bot to enhance airport operations

The optimization system

This project aimed to provide airports with a powerful tool for more efficient planning and the assignment of aircraft to stands—essentially the equivalent of gates. This partially rule-based system was designed to follow established constraints while also optimizing operations based on the business value these stands could generate.

Objective

The primary goal was to enhance airport operations by providing a tool for efficient aircraft assignment to stands. This involved utilizing both rigid rules and business KPIs to maximize revenue generation and improve passenger experience.

Challenges

Key challenges involved balancing resource placement to support revenue opportunities—such as encouraging traffic through high-value areas—while also ensuring that transitions within the hub were smooth and timely for connecting passengers. Additionally, sustainability goals required the system to reduce environmental impacts by minimizing travel distances and fuel consumption through optimized placement strategies.

Approach

The tool employs a dual approach that integrates predefined operational rules with business-oriented KPIs to support dynamic optimization. To inform its strategies, the system analyzes also historical data to pinpoint areas with high revenue potential, allowing for optimal resource placement.