Get AI-ready with Microsoft and HSO
Machine Learning for better pricing proposals
After a certain number of flight hours it is time to change oil, tires and other aircraft parts. Airlines have maintenance contracts for that. For example with a party like Spairliners. But how do you determine realistic prices in these contracts? A challenge in which HSO helps them. Head of Innovation, Data & Quality at Spairliners Jonathan Mayer: “HSO helps us to gain the insights needed to determine the right rates using data and machine learning.
Before the arrival of HSO, Spairliners was constantly calculating, gambling and estimating what a maintenance contract should cost for an airline. Spairliners based the price on the number of flight hours and estimated the required maintenance on that basis. This estimate was not accurate due to incomplete data. As a result, there was no unambiguous pricing policy and it affected planning and inventory management. “Because we had good experiences with HSO when implementing a CRM system, and they are specialized in Data & Analytics, this was the time to fly them in for this issue,” says Jonathan.
From strategy to dashboard
“Together with HSO, we have started making an inventory of the large amount of data in all systems and looking at how we can make better use of it. The goal: to realize a data-driven working method, with which we can accurately determine contract prices, among other things.”
Getting the data ready for reliable outcomes
Irene Reijntjes, data science consultant at HSO, adds how they assess the quality of the data: “You can only make reliable predictions with high-quality data. HSO has developed a method to test and specifically optimize the data landscape and quality. For example, by checking whether there are no duplicate records and whether there is consistency in values within a dataset or between different datasets. ”
After assessing the data, the team could move on to realizing the architecture, dashboards and reports. Irene explains what exactly they made for Spairliners: “The solution we created was made in Databricks. The only thing that Spairliners needs to do now is to prepare input files in the correct format, consisting of, among other things, fleet specifications.
This input is processed in Databricks. Ultimately, we use Machine Learning to predict the expected maintenance and associated costs. The results are stored in a data warehouse and visualized with Power BI. This gives the user quick and clear insight into the amount of maintenance required for a specific airline and associated costs.
Get AI-ready with Microsoft and HSO
“Thanks to HSO's solution, we are not kidding ourselves with prices that are too low, nor are we asking too much of our customers.”
Planning and optimizing
Jonathan sees HSO's solution as an enrichment of Spairliners' business model. “We now have a new working method, which means that we know that our prices are substantiated. That feels good, because we never ask for too little. And we can explain it more easily to the customer. HSO listens to what our team needs and sets to work. In this way they continuously optimize the tool according to our wishes.”
Irene sees plenty of opportunities for that optimization to help Spairliners move forward. “The data used has been influenced by corona. The models that predict how much maintenance will take place are based on historical data. Due to corona, flight hours were lower than before and we saw this reflected in the data. The patterns in the data have changed significantly as a result and it is therefore important to adapt the solution accordingly. We are happy to take the initiative to see how we do that with Spairliners.”
Customer stories on Data, AI and Machine Learning
Get in touch with our Data & AI experts