Airport Capacity Planning Data Scientist

You will develop and maintain machine learning models, simulation tools, and analytical frameworks that directly inform capacity planning decisions across one of the world's busiest airport baggage systems-processing in excess of 80 million bags annually across four terminals.

Your responsibilities and activities will include:

• Design, build, and validate machine learning models (e.g. XGBoost, GBM, random forests) for forecasting baggage volumes, capacity utilisation, and recirculation rates across all terminals.

• Engineer temporal and lag-based features from high-volume trace datasets (180M+ historical rows, 80,000+ daily ingestions) to improve model accuracy.

• Conduct hyperparameter tuning, cross-validation, and model selection using rigorous statistical methods; target production-grade performance metrics (e.g. R² > 0.95, low RMSE/MAE).

• Develop time-series decomposition and anomaly detection pipelines to identify emerging operational bottlenecks.

• 2.2 Simulation & Capacity Analysis

• Build and calibrate discrete event simulation (DES) models of terminal baggage systems to stress-test capacity under various demand scenarios.

• Produce peak-flow analyses, what-if modelling, and scenario planning outputs to support infrastructure investment decisions and airline schedule changes.

• Translate complex analytical outputs into clear, actionable capacity recommendations for operational stakeholders and airline partners.

• Develop interactive Shiny applications and dashboards for real-time and historical performance monitoring.

• Create publication-quality reports and presentations using LaTeX and PowerPoint for senior leadership, airline customers, and Heathrow Airport Ltd stakeholders.

• Present findings to non-technical audiences, distilling complex statistical concepts into clear operational insights.

• Stay current with advances in applied machine learning, operations research, and airport technology.

• Identify opportunities to apply AI/ML techniques (e.g. GPU-accelerated training, LLM-assisted analysis) to improve operational decision-making.

• Contribute to the team's code standards, documentation, and reproducible research practices., * YuLife - Wellbeing membership with fast access to GP appointments, promotion of health and wellbeing along with daily quests to gain Yucoins that can be swapped for shopping vouchers

• A challenging work environment with lots of opportunities of career progression.

• Cycle to work scheme

• Yellow Nest is a salary exchange scheme that reduces childcare costs for parents and employers

• Pension with Aviva

• Access to Achievers an award-winning recognition platform that inspires to recognise your coworkers Where points are awarded that can be exchanged for a range of goods and discounts.

• Master's degree (or equivalent) in Data Science, Statistics, Operations Research, Computer Science, Mathematics, or a closely related quantitative discipline

• Demonstrable portfolio of applied machine learning projects with real-world datasets

• Advanced proficiency in R programming, including tidyverse, data.table, caret/tidymodels, xgboost, and Shiny

• Strong SQL skills with experience querying large-scale relational databases (Azure SQL, PostgreSQL, or equivalent)

• Hands-on experience with DuckDB or similar columnar/analytical databases for high-performance local analytics

• Solid understanding of supervised learning algorithms (gradient boosting, ensemble methods, regularised regression) with practical deployment experience

• Experience with feature engineering for time-series and operational data, including lag features, rolling aggregates, and temporal encoding

• Proficiency in data pipeline development using Azure Data Factory, or similar orchestration tools

• Proven ability to translate business problems into analytical frameworks and deliver actionable recommendations

• Excellent written and verbal communication skills; comfortable presenting to senior leadership and external stakeholders

• Strong problem-solving mindset with attention to statistical rigour and reproducibility

• Ability to work effectively within a team of 9 analysts while managing independent workstreams

• Experience in aviation, airport operations, logistics, or baggage handling systems

• Familiarity with discrete event simulation tools and methodologies

• Knowledge of LaTeX for technical documentation and report generation

• Experience with GPU-accelerated machine learning (CUDA, cuML) or high-performance computing environments

• Exposure to version control (Git) and collaborative development workflows

• Familiarity with Python as a secondary language for interoperability