Local Ocean

Saving sea turtles in Kenya using AI

When local communities are involved in conservation efforts, those efforts have a far greater chance of success. In 1997, a group of residents from Watamu, Kenya, decided to take action against increasing human-related threats to the marine environment in their neighbourhood. They were particularly concerned about nesting sea turtles, who were unable to successfully lay their eggs due to beach developments and the risk of poaching. A dedicated team from the community patrolled the beaches day and night to protect the female sea turtles coming to nest, their eggs and their hatchlings. To date more than 1600 nests have been monitored, and more than 157,000 hatchlings have made their way to the ocean.

In 2002, recognizing the wider need for community involvement in marine conservation, the initiative was renamed Local Ocean Conservation (LOC) to encourage people to look after their local marine environment. LOC recently went on a significant digital transformation journey, consolidating turtle records from multiple separate offline databases to facilitate data sharing and analysis.

The primary goal for Local Ocean is to use AI to develop their image recognition and data-capturing tools to develop capability that can be used in the field to identify sea turtles, doing away with the need for invasive and stress-inducing metal tagging. Although metal tagging is currently best practice in the marine conservation sector the current tagging system is also expensive, which puts it out of reach for smaller conservation groups to use. This limits the amount of important turtle data collected within the sector, and particularly in the West Indian Ocean, which has a direct impact on how conservationists in the area can effectively monitor and model the turtle populations and key data on a per-turtle basis.

Partnering with Microsoft through the 4Afrika initiative, Local Ocean now has a single, unified cloud-based database with more robust data collection, using Microsoft Azure, which helps store data and enables data visualisations via Microsoft Power BI.

“What’s really important to mention here is the technology assessment we did before choosing our preferred technology platform, with Microsoft coming out as the preferred option. This was mainly due to the concern that we have for technical assistance and skills to maintain what we create. Kenya has a deep talent pool for Microsoft support,” says Local Ocean CEO Justin Beswick.

LOC deployed Power Apps to create the first Beach Sustainability map in Kenya and doing so at a very low cost – a significant achievement for the team. As little was known about turtle nesting sites in Kenya, the team conducted a baseline assessment of the suitability of beaches for turtle nesting along the Kenyan coast. To enable this exercise, they developed a mobile app that allowed them to assess nesting suitability and mapping of socio-economic activities. This identified human-induced factors that contributed to a beach being categorised as unsuitable for nesting. During the exercise the team mapped a total beach length of 78.8km. This allowed LOC and stakeholders, for the first time, to have a clear idea of the suitability of Kenya’s beaches for turtle nesting and identify where urgent conservation efforts are required. From the analysis, 17% of the beach was found to be unsuitable due to human activities influence, which is greater than the 13% that is suitable for nesting.

Technology can be a huge enabler to conservation if deployed correctly. Local Ocean hopes to develop effective tools for marine conservation efforts that can be used not only to boost their own efforts, but to benefit others, with a particular focus on the African continent.

Related articles