Getting the science we need for the future we want

For me, one of the best things about working in data and AI, especially at Capgemini, is that we are constantly looking for ways to solve problems – both for organizations and society at large.

In fact, each year Capgemini hosts a companywide hackathon, the Global Data Science Challenge (GDSC), where employees compete to build AI solutions to help solve real-world challenges.

Last year we teamed up with marine biologist Lisa Steiner, renowned for her sperm whale research in the Azores. She tracks and monitors whales, identified by their tail fins, to learn about their migration routes, social structures, and to protect their natural habitats.

But this requires manually identifying whales by matching new whale pictures with historical ones – tedious and labor-intensive. But not anymore!

The winning GDSC team built an AI solution, trained using Amazon Web Services (AWS) machine learning and 4,500 tail-fin images from more than 2,200 whales. Now, Lisa is freed up for more important tasks, contributing to amazing work protecting the sperm whale, listed as vulnerable on the International Union for Conservation of Nature (IUCN) list.

We’ve since been integrating this into Happywhale, helping to crowdsource humpback whale photos to provide more valuable data. The Happywhale team are now extending their reach beyond humpbacks, and we’re glad to support their mission.

The challenge for 2021: LoVe the ocean

For this year’s GDSC, the fourth edition, Capgemini employees were given an enormous task: to help manage ocean data and contribute to ocean conservation. Specifically, they were asked to assist the Lofoten-Vesterålen (LoVe) Ocean Observatory, located on the coast of Norway’s Lofoten archipelago.

LoVe uses a system consisting of fiber optic subsea cables and sensor nodes that cover a large area from land to deep sea to gather and stream real-time data about the local ocean environment.

This data is used to understand the overall state of the ocean, predict the variability in the ocean environment, and study human impact. But the volume of data was becoming unmanageable. Over the past decade, LoVe has collected over 100TB of data… and it only continues to grow! You can imagine how difficult it is to manually assess and evaluate it, let alone analyze and identify relevant insights for scientific research.

That’s why LoVe urgently needed new automated methods to help them in their work. Luckily, Capgemini teams were up to the challenge.

Unleashing human energy through technology

This year, 673 teams – comprising 1,200 Capgemini colleagues – competed to develop a concept for an AI-based solution to LoVe’s problem. And two teams emerged victorious to share the top prize of seeing their solutions in action: one based in India and the other in the UK. The winning solutions centered on processing data in real-time and automatically extracting anomalies in the datasets, again using AWS ML services.

Here’s how the teams approached their solutions:

“We tackled each data source individually, designing a model that would identify the outliers in each set. Therefore, much of our focus was on the data pre-processing, to determine the most relevant variables among the thousands we were presented with,” says Anupam Saha, senior delivery manager at Capgemini in India and leader of one of the winning teams.

It was a similar situation for the UK team, led by David Gilhooley, principal consultant and engagement manager, at Capgemini in the UK. “We had to take these multiple data sources and organize them day by day,” explains David. “As well as padding out the missing areas, we had to normalize the data in order to deploy the machine-learning analysis correctly.”

These incredible solutions are already adding value: there are several areas that the teams picked up that LoVe is now exploring – like whale vocalization, oceanographic and tidal features that hadn’t yet been recognized, and the strong influence of weather on the environment.

Save the data, save the world

It’s always amazing to see the work we do have a larger and real impact, which goes way beyond the day-to-day data insights. By assisting the LoVe team, we are helping to conserve a key part of our global ecosystem – the ocean. We are contributing to its protection, but also to our collective understanding on how it can act as a solution to combatting climate change.

I’m incredibly proud of all the teams that participated in this year’s GDSC, and to be a part of a company that actively seeks to challenge ourselves to get the future we want with technology. I will be following the outcomes of these exciting solutions and can’t wait for next year’s challenge!