Rolls-Royce Data Challenge 2019, Singapore

Rolls-Royce rolled out the Data Innovation Challenge – Asia looking for solutions that can leverage on flight data being collected every day to solve significant challenges in the aviation industry.
4 primary data challenges were introduced by Rolls-Royce to enthuse startups with strong data & Artificial Intelligence capabilities to help address these challenges.

Navigem data picked on 1 such challenge- “How might we optimize flight schedules that take into consideration factors such as weather and congestion?”

Towards that a summary model incorporating the influence of the factors like delay propagation, passenger delay, carrier delay, & weather as a function of Airfield throughput would contribute to predicted precise arrival times and thus provide for near real-time dynamic scheduling across the trip time, contributing to reduce Delay Propagation.

As pure practitioners of Big Data, Analytics and Deep Learning technologies (Machine Learning, Synthetic Intelligence) and Data Sciences, we at Navigem proposed a Data-driven Decision Support System to optimize the flight scheduling. The Pure data-play approach is built on the fulcrum of availability of rich and diverse data sets related to various facets of Air Traffic Management at origin & destination, like, but not limited to, silos–

• Aircraft-specific info

• Weather Info (Origin & Destination, Enroute based on long-term averages & empirical observations )

• Aggregated Delay (if any)

• Congestion Levels (based on arrival time estimates & historical data at

• Origin & Destination, qualified by arrival & departure)

Pertinent quality data subjected to Data Sciences (Correlation, Regression, Random Forest Regression, Gradient Boosting Algorithm and such) led  to a well-defined training Dataset, which ensured a constant, consistent, regular optimization of the ever-evolving Machine Learning algorithm,  continually enhancing the precision of the Arrival Prediction, thus facilitating  congestion mapping & handling and, by extension, effective Flight Schedules.

Team Navigem participated in the data challenge at Singapore and was given the opportunity to present its solution for one of the themed problem statement.

Going forward, Navigem Data is in talks with R2 Data Labs to collaborate with RR and work on flight challenges purely through data!