Sonal Ambwani: Algorithms of the future

by Sonal Ambwani on 02/03/2020

My time at L3Harris has been one of continuous learning by working with some of the smartest people in the industry. The work we do is at the leading edge of Artificial Intelligence and deep learning innovation.

In broad strokes, my team and I ensure that the process of screening passengers through the ProVision body scanner is as secure and comfortable as possible. The scanner is an integral part of the larger suite of L3Harris’ aviation security devices and solutions. It is responsible for screening passengers for any threat objects they might be carrying on their person. It uses millimetre wave technology to image passengers. These rays reflect differently with and without the presence of specific threat objects or anomalies and therefore presents unique information that make detection of such objects possible via algorithms. That’s where my team and I step in.

My responsibilities are to create and deploy artificial intelligence driven threat detection algorithms on the images formed from the millimetre wave data. These algorithms are responsible for detecting objects not allowed aboard an airplane, while also maintaining a steady throughput through the security checkpoint. In other words, they are fast and accurate. We start with collecting and curating ProVision image data that will best suit the purpose of enhancing threat detection and lowering false alarms. Data collection is followed by labelling and annotating data using our special labelling tools. Once we have enough labelled data available for our deep learning algorithms, we enter the stage of development. The good news is that over the last couple of years we have honed an internal AI development framework that allows us to be highly innovative and efficient. Our infrastructure allows us to quickly develop, test and deploy our algorithms and therein lies the key to our success.

The ProVision system in its current iteration provides a secure, fast and discreet way of screening passengers at security checkpoints in airports throughout the world. My team and I continuously strive to make the current and future versions of the system even better in achieving the goal of safer skies. There has been a fair share of challenges and rewards. Overall, it makes coming to work every day a great deal of fun and a big challenge.

I have a greater appreciation of our commitment towards continuously raising the standards of airport and passenger security. As an employee, it is very encouraging to see the willingness towards adopting the latest AI technology allowing us to advance the industry forward.