Aksha Srivastava: “Artificial Intelligence has a big role to play in the Aerospace domain”
What made you so fascinated with the field of Artificial Intelligence?
A.S.: As a computer science engineer it is in my nature to solve problems, and when I came across AI – a technology that makes it easier to solve complex problems and simulate human behaviour, especially when it comes to logical decision-making and predicting values - it only made sense to take that up as my research area.
In light of the current technological boom, immense amounts of data are being generated. The fact that this data can be used to make decisions and detect patterns intrigues me greatly.
What drew you towards the Aerospace industry?
A.S.: My father is an Indian Air Force officer, so I have an inherent love of Aerospace. After my Amazon experience, I really wanted to implement the knowledge I had gained in a real-time environment in the Aerospace domain, and Safran seemed like an obvious choice!
When we think about Artificial Intelligence and Aerospace independently, they are far from interrelated; what are some applications of AI in the Aerospace industry that you can imagine being implemented today?
A.S.: The data from all branches of the Aerospace industry, be it design, manufacturing or operations, is produced in immense quantities, which results in AI having a variety of applications in the field.
For example, Predictive Maintenance and Flight Data Analytics can make recurring patterns evident and manual checking and use of human intuition can be replaced by Deep Neural Networks (DNNs).
Applying AI on this can enhance the design capabilities, increase production rate and also result in more efficient operations.
Artificial Intelligence can also cut down the cost of manufacturing by observing facilities and taking insights from what people learn over time and incorporate it into systems to avoid skill loss during knowledge transfer.
In the immediate future, we will be eyeing IVHM technology to replace some of the existing aircraft systems and also add a layer of performance for improving functionalities.
What are some projects you have done or are currently working on?
A.S.: When I joined the team, I was helping the product support engineering team to develop the concept of an expert system for Symptom Fault Correlation, for an Indian customer working on the Defence Helicopter programme, as a part of the CBT programme (Computer-Based Training).
We're currently working on providing an innovative, out-of-the-box idea to perform the Concessions treatment process in Safran Engineering Services India. The project is called Concelligence, which we thought was a witty combination of Concessions and Intelligence!
The goal is to come up with an assistive algorithm by learning from all the previous concessions and results and drawing on efficient engineering programmes to carry out the treatment using AI/ML techniques and progressively reduce the time spent on this non-value added task!