How Akila Is Optimizing Train Station Efficiency With Digital Twins and AI Agents

Creating digital twin solutions for complex, dynamic public spaces like train stations can be challenging. It requires real-time integration and rapid analysis of vast, diverse data streams from multiple systems for safety and efficiency, while taking into account traveler experiences. Simulating logistically challenging environments is far more demanding than residential buildings, including factors like unpredictable crowd surges and geographical elements such as underground terminals.

Akila’s digital twin platform brings physical AI to smart cities, following the NVIDIA Blueprint for smart city AI. It uses live and recorded visual data to centralize building analytics for complex environments, such as rail terminals, universities, and other types of real estate. By integrating digital twins, sensor data, and advanced simulation, Akila visualizes all building information, enabling smarter, more efficient, and sustainable operations.

French train network operator SNCF Gares&Connexions is working with Akila to optimize and future-proof its network of train stations and railways across France and Monaco—operating nearly 14,000 trains daily.

SNCF Gares&Connexions simulates several processes to ensure that all safety protocols are functioning correctly and to control the flow of passengers, staff, and trains. With a digital twin, it can also:

• Simulate and analyze the impact of future building refurbishments
• Analyze airflow, temperature, and daylight simulations for optimization of heating and cooling systems 
• Train future robotic or autonomous devices in a digital, physically accurate learning environment
• Simulate emergency scenarios, such as evacuations

Having a digital twin of a station enhances operational efficiency by providing better situational awareness and real-time alerts. To make decisions quicker and easier, the digital twin can proactively flag potential issues such as water leaks or abnormal energy consumption, before they become major problems.

Akila’s platform also supports sustainability efforts by reducing energy consumption and identifying areas for improvement, such as simulating airflow sequencing. By analyzing different scenarios using real-time data and insights, SNCF Gares&Connexions can make informed decisions about energy- and cost-efficient practices and asset upgrades, resulting in a more sustainable and resilient infrastructure.

Early versions of Akila’s digital twins faced common struggles with model expressiveness, data quality, scalability, and simulation accuracy. This limited their usefulness for advanced analytics and timely insights in large buildings and spaces.

The combination of NVIDIA Omniverse and OpenUSD helped solve these limitations by letting Akila create physically accurate digital replicas and run real-time simulations. Now, Akila can simulate complex 3D scenes such as solar heating systems, air flow, and crowd movement in a digital twin. With NVIDIA Metropolis, a platform for vision AI application development, Akila aggregates and streams thousands of live camera feeds, recorded video, sensor data points, and building management systems ported directly into their platform. These advanced capabilities are crucial for understanding and optimizing the performance of complex spaces, like transportation hubs and corporate buildings. 

SNCF Gares&Connexions uses NVIDIA RTX™ graphics cards for local on-premise processing and NVIDIA A10 GPUs in the cloud on Microsoft Azure to ensure scalable and high-performance graphics. This combination enables real-time simulation and data integration, plus a fully sovereign platform that complies with local data regulations and specific security requirements.

Simulating SNCF Gares&Connexions train terminals and collecting building analytics has delivered significant operational and environmental benefits:

• 50% reduction in downtime: The platform's real-time monitoring and alerts have significantly reduced the time it takes to identify and address issues.
• 100% completion of preventive maintenance on time: By providing advanced insights, the platform helps ensure that all preventive maintenance tasks are completed on time, improving the reliability and longevity of the facilities.
• 50% reduction in response times to issues: Real-time alerts and situational awareness have cut response times to issues in half, allowing for faster and more effective problem resolution.
• 20% reduction in energy consumption: The platform's energy management capabilities enable simulation of potential upgrades and modifications to optimize systems and assets, resulting in a 20% reduction in energy consumption.
• Rapid intervention: The platform saves operators two hours per week by providing real-time data and digital twin simulations, simplifying decision-making and resource allocation.

Akila’s solution has opened up new use cases, such as safety and security applications, which are vital for large spaces like retail malls, transportation hubs, sports stadiums, and campuses. These new capabilities allow businesses to enhance their operational resilience and transparency, leading to better decision-making and resource allocation.