Smart Inventory Management & Optimization

Team members

Chester Lim Zi Hao (ESD), Cui Wenqi (ESD), Hour Youlinserey Devid (ESD), Chia Yu Ying (ESD), Chen Ken (ISTD), Sun Kairan (ISTD), Justin Peng Zheng Wei (ISTD)

Instructors:

Francisco Benita, Cyrille Pierre Joseph Jegourel

Writing Instructors:

Susan Wong

Teaching Assistant:

Lim Swee Hao

Did You Know?

Rotables are expensive parts of an aircraft. Each part can cost up to USD1 million.

Problem

Traditionally, the manual way of handling thousands of rotable inventories with tons of messy spreadsheets is time-consuming and prone to errors. This incurs huge amounts of unnecessary cost for Singapore Airlines Engineering Company (SIAEC). As inventories increases in volume to meet the growing demands in the aviation industry, a more efficient way of managing inventories is vital to ensure high service levels at minimal cost.

Introducing

4 Key Features of SIMO ↓

Concrete futureproofing on
rotables provisioning

Rotables vary in prices and importance to the aircrafts. Inventory optimization is vital to ensure that inventory costs are minimised while maximising service levels. By formulating an optimization model and solving the optimization model using the SCIP solver helps SIAEC to better provision for rotables. An in built-optimization algorithm integrated with a dynamic parameter-setting page in SIMO allows user to simulate real-life scenarios, assess potential risks, and make better purchasing and stocking decisions.

Dynamic models to forecast
rotable's at a moment's notice

Most rotables are bulky and hard to transport. Rotables not in used are kept in the cold room to slow their rate of deterioration, and they are only taken out when needed. Hence, forecasting the demand of rotables is vital to ensure that rotables are transported only when there is a need for it in order to minimise costs and maximise efficiency. Our forecasting engine includes state-of-the-art forecasting models such as NeuralProphet and Prohet which are auto-trained using historical data to predict future demands. Forecasted results from the best model with the lowest Root Mean Squared Error will be auto-displayed to provide insightful recommendations to users.

State-of-the-art modelling for
market trending prices

This feature utilises a combination of historical tracking of prices between each category components and the addition of state-of-the-art forecasting methods. Some methods includes the Autoregressive Integrated Moving Average (ARIMA), Holt Winters Method and Neural Prophet. These models helps SIAEC to track the historical trends of market prices across vendors and assist SIAEC to make informed choices by refreshing the forecasted prices for each category of components.

Rotables Performance made
available and automated

This component of the project helps SIAEC understand the in and outs clearly of their historical performance tracking against international Original Equipment Manufacturer (OEM) benchmarking, as well as the average performance performed globally around the world. On top of that, having an automated underperformance trending feature benefits SIAEC whereby it allows them to put their focus on key rotables that are constantly underperforming relative to global standards.