Yieldr and Owlin have been approved for an MIT research & development grant from the Dutch Government (Rijksdienst voor Ondernemend Nederland).
As stipulated by the application process, Yieldr partnered with fellow Dutch data science company Owlin and together they were granted the maximum amount to conduct further research and develop forecasting tools to make airlines more sustainable. This affirmation from the Dutch Government is further validation of Yieldr's product vision.
“We’re building our product with a clear purpose to facilitate efficient, sustainable and profitable air travel growth,” said Yieldr founder Mendel Senf.
“Approval of this MIT grant further validates that our vision is solving a clear environmental concern with added economic benefits, both for the short and long term.”
Yieldr's forecasting capabilities will be extensively improved by incorporating Owlin's data and will allow airlines to better understand how news is influencing their demand. This system will predict the demand for airline tickets on different routes and at specific times, based on data sources other than just historical data.
Owlin’s data is already used in forecasting counter-party risk and settlement risk for portfolio managers. Using its technology to better predict airline tickets is an interesting use case where their knowledge of financial models already creates a head start.
Owlin brings their experience in data science and machine learning to understand how news items impact/informs financial decision making. For airline forecasting, they will analyze destination news at scale and establish how that news can impact demand.
Replacing historic data with other data sources for forecasting is an innovative approach for the airline industry. Even without modifications to the demand forecast itself, integration of external data sources such as Skyscanner metasearch data and Predict HQ event data are powerful tools for airlines to make better decisions, as these other sources can show the reason behind the flow of demand curves.
Most often revenue managers have a vague idea of why demand is being affected since they already take some sort of external data into account, but do not do so in a complete and integrated manner. Harmonizing all the factors that influence demand in one interface can greatly improve the development of the intuition needed to take decisions.
This solution adds economic value by increasing per-flight revenue and bookings, with its greatest benefit being its environmental impact. Namely, the service will lead to a decrease in CO2 emissions per passenger kilometer, and lower waste, noise and air pollution by getting more passengers to fly on fewer planes.
In today’s economic and environmental climate, sustainability has become a major initiative within the airline industry. Flights produced 859 million tonnes of CO2 in 2017, which is 2% of all human-induced CO2 emissions. More than ever before, airlines are striving to be more sustainable.
The proposed service of this MIT R&D project contributes to this goal by increasing load factors and thereby reducing CO2 emissions per passenger kilometer and potentially even the number of flights.
Although the airline industry has not been included in the Paris agreement, the industry aims to regulate itself in anticipation of EU-issued policies. One of the main goals is carbon- neutral growth by installing a cap on net aviation CO2 emissions from 2020. In order to achieve this target, a four-part strategy has been launched. One of the pillars is “more efficient aircraft operations.”
Air passenger growth is growing exponentially. The International Air Transport Association (IATA) projects air travel growth doubling to over 8 billion passengers by 2035. At the current rate without action, airlines will account for 27% of all human-induced CO2 emissions by 2050.
According to IATA, 81.5% of seats were sold globally in 2017. If airlines can better manage their demand, they can increase load factors and eliminate waste by getting more passengers to fly on fewer planes.
The aim of the new service is to increase this load factor by 5%. This is to be achieved through the innovative combination of (local) macroscopic and (individual, anonymized) microscopic Big Data sources, and applying AI and Machine Learning technology.
The Yieldr-Owlin proposal was one of 40 applications (ranking 23rd overall) accepted from a pool of 130 eligible applicants, from which Rijksdienst voor Ondernemend Nederland had a budget of 8.4 million EUR to disperse.