To realize the full power of machine learning, organizations need to focus on operations and delivery of the predictions as much as the development of the algorithm itself. In the case of NASA and ATD-2, not being able to trust the projections generated could be a life-or-death situation. Air traffic controllers must trust the recommendations presented by them, and validation is essential towards building trust.
Mosaic sought to provide an open and accessible model to predict the impacts of wind on air miles flown under an SBIR contract with NASA. Ultimately, this effort has developed a model that the broader aviation community can use as a publicly available data service.
Mosaic ATM Case Study | ATD-2 Addressing Inefficiency in Today’s Air Transportation System According to NASA1, many of today’s air transportation system issues can be attributed to a lack of information sharing amongst the operators responsible for managing air traffic in busy terminal environments. Concepts and technologies to improve the Read more…
As part of the ATD-2 effort, Mosaic, in conjunction with NASA, evaluated an ML-driven approach against the controller-derived assignments, hypothesizing that the performance would be equivalent to that of an expert air traffic controller.
We decided to approach this problem as a similarity learning modeling effort. We used convolutional neural networks to train a model that takes an image or video as input and outputs a vector representation of the input, such that similar inputs will be close to each other in the vector space. The vector learning is driven by a triplet loss function.
Having an autonomous artificial intelligence (AI) system that can monitor individuals via facial mood recognition, vocal tonality analysis, proximity to one another, performance, biosensors, surveys, and more, and predict conflict before it is problematic could improve a unit’s cohesion and performance in missions both in space and in isolated environments on Earth.
This white paper explores how traditional models of the value of information (VoI) can be extended effectively to account for uncertainties presently inherent in gathering and analyzing big data. To illustrate the challenge, we explore the VoI an automobile manufacturer may derive by engineering a telematics system into its vehicles.
Working in conjunction with subject matter experts, data scientists can swiftly apply statistical tools and uncover emerging trends. This is extremely valuable for companies trying to operate in a disruption. Not only will executives have an accurate representation of their present situation, but new products & services can be devised from these insights.
Designing and deploying machine vision is a powerful technology that humans can employ to improve their decision making. These deep learning and AI techniques are not easily developed, and trained data scientists need to be involved in the translation of analytics to business insights.