From big-rock problems to drawing interconnections: The Game Changing Story of An Airliner that took flight


  • THE MU SIGMA TIMES
  • August 6th, 2019
  •   1774 Views

Today, if I were to fly (read Airline) from Bangalore to any destination, international or domestic, what would decide my choice of service? Cost? Time to travel? Number of transits? Location of Transit? Schedule? Past experience? In-flight experience? On-ground experience? Well, I can keep rambling the questions, because I am a fussy customer to please.

Now, imagine millions of customers like me for the airline service provider – fussy, over expecting, critical of every small good or gap. It is no easy joke. This behavior of mine is given further impetus by the hyper-competition in the sector today – budget airlines, customized service offer, ability to cross sell and up sell in real time, air miles and so on.

One of the major reliability measures for Airlines today is On Time Performance – it is a key differentiator. And while it sounds simple, the imminent complexity of this measure is a nightmare of many big Airlines brand. So then what can Brands do to stay ahead of the race and make the cut – for customers, the stakeholders, and the business itself?

This is our engagement story with one of the largest low-cost Airline in the US – in an interview with Anubhuti Sood and Suhani Jain of the Account.

Humble beginnings

“Mu Sigma and the client had been engaged for a while. The long standing relation was founded in our mutual belief in the interconnected approach to problem solving. This was put to test when the VP of strategic planning approached us with a challenge – he was keen to understand how the decisions made in the last 10 years impacted the business model of today. The client used aircraft utilization as a metric to measure the productivity of aircrafts. Our initial reading of the numbers pointed to the fact that at the cost of maintaining high Aircraft Utilization, there was a steep decline in OTP. We had to find a way to balance the AU and OTP,” said Anubhuti Sood.

A discovery driven approach

We followed a two phased approach to solve the problem of declining OTP for the client: Hypothesis driven approach followed by simulation modelling.

This approach was taken to make sure we had absolutely explored all possibilities related to OTP, at the same time testing and learning from a simulated and prototyped environment.

Here is where the irreverence played up. The fact that OTP had multiple moving parts that were interconnected required a solution that could accommodate the complexity of the connections while keeping it simple and scalable – the art and science had to come together. Using simulation (where advanced Math and Technology come together) we could overcome complexities. Add to this design thinking and business acumen, we were able to build an accurate solution. This model treated every route-month and station-time of day-month separately, thus encapsulating the complexities implicitly within the solution itself.

Towards a game change

Measure of On-Time Performance (OTP) impacts Customer Satisfaction. It is a unifying metric that is the sum total of efforts put in by various departments. Accurately predicting this metric becomes all the more important because under predictions may lead to extra cost due to incentives, while over prediction may lead to underperformance and hence Customer dissatisfaction.

“Using Simulation we were able to get an accuracy of 98.5%,” – Suhani Jain

Additionally we were able to: provide the client one version of the truth, as opposed to different departments providing their own forecasts; align and work together with the leadership team of the client to achieve the common organizational goals; the model was used for long term business decision making and solving interconnected problems such as Fleet planning, operations and resource management, Schedule Planning, Cost Predictions, Revenue Investments, and Customer Satisfaction.

Against the norm

But none of this came without its share of resistance and friction. These were to be attributed to two major reasons: one, the approach of using Simulation was black box to them – it was a new technique; and second, different teams in the client ecosystem were already working on predicting OTP and the Corporate Strategy stepping into their territory was not very welcome.

Resting the case

All apprehensions were allayed – the results spoke volumes. This accomplishment was filled with its share of best practices. We were able to:

  • Educate the clients: In keeping with our vision of wanting to institutionalize data driven decision making, we ensured the client was included in every step of the problem solving process. We collaboratively defined the problem using muPDNA, oriented the client towards the complex mathematics that formed part of our solutions, and laid out the next steps by understanding the inter-connections in their ecosystem.
  • Maintain continuous feedback and scale: We continuously re-visited the simulation framework to make it better and more accurate. The process also helped us identify opportunities of automation in R.
  • Translate our relation from vendor partners to trusted advisor: Despite the challenges, we were able to sensitize our clients into the importance of institutionalizing analytics for informed decision making. This was made visible to them by drawing the interconnection, providing necessary orientation, helping them deliver results in an informed manner.
  • Popularize the Mu Sigma Artifacts such as muPDNA and muUniverse in the client ecosystem.

There is more to explore

Airlines, as in other industries, is shifting focus from providing low cost travel to providing better customer experience. Analytical study of lean operational structure is enhanced through systems thinking and simulations techniques.

Today, the consolidated efforts that we put has paid off – the client has implemented the simulation framework across their organization to make informed business decisions.

On the whole, simulation can be used to duplicate features, appearance and characteristics of a real system. It could also double up as a sandbox for trial and experimentation. The beauty of this technique is that it can be used in almost all fields and across domains, be it pricing policies, forecasting, marketing strategies, stock and commodities analysis, and determining the processing time for every step of the manufacturing cycle. Be it training, process improvement, predicting outcomes, or managing risks, simulation is the way forward for organizations that wants to continuously innovate at low cost and low risk, and stay ahead of the race.