Across the savannahs of Laikipia, Kenya, people, livestock, and wildlife share a landscape that is both rich and contested. Elephants uproot crops in a single night, large carnivores prey on livestock, and these conflicts can lead to lethal control of predators, threatening both human livelihoods and biodiversity. Yet, coexistence is possible if innovative approaches can reduce the costs of sharing space with wildlife.
Since June 2024, we have been developing and piloting Ambiance, a novel autonomous playback system (APS) designed to mitigate human-wildlife conflict in multiuse landscapes. Ambiance is a solar-powered device capable of broadcasting cue-specific acoustic treatments over extended periods, creating the illusion of human presence to deter species that raid crops or attack livestock. This concept builds on long-standing ideas in behavioral ecology, translating them into a conservation tool with direct, on-the-ground impact.
Our pilot experiment at Mpala Research Centre tested the potential of Ambiance at an unprecedented scale, deploying 50 devices across two 1-km² grids over five months. Each device broadcast human conversation or bird vocalizations for ten hours daily, while camera traps, artificial forage stations, and vegetation surveys recorded responses. The results were striking: elephants reduced space use by 51%, activity by 66%, foraging by 80%, and tree damage by 64%. Mesocarnivores and rodents also altered their foraging behavior, and these effects persisted for weeks after broadcasts ceased — demonstrating that simulated human disturbance can reshape wildlife behavior over extended periods.
These findings revealed both the ecological power of acoustic playback and the potential for Ambiance to transform human-wildlife coexistence. Importantly, they showed that even calm human voices could deter conflict species, and that fine-tuning tone, language, volume, and stimulus type could enhance efficacy. They also sparked strong interest from conservation partners, including SANParks, the Amboseli Trust for Elephants, the Kenya Wildlife Service, and local community conservancies.
With WILDLABS and Arm support, we will build on this proof of concept to develop a third-generation Ambiance unit, optimized for usability and performance, and establish a Kenyan-based assembly process. We will manufacture 60 devices and deploy them in a year-long randomized controlled trial with pastoralist and agrarian households in Laikipia. Participating households will host Ambiance units broadcasting either human disturbance or bird vocalizations, with treatments rotated systematically to assess deterrence effects on elephants and carnivores. We will integrate community-sourced recordings to ensure cultural and ecological relevance and measure outcomes with standardized reporting.
This project is both ambitious and feasible. It leverages over 50 years of conservation experience in Laikipia, a proven APS design, and strong local partnerships. By the end of the award period, we will have a fully functional Kenyan Ambiance assembly process, a tested and scalable technology, detailed open-source documentation, and a robust evidence base for APS as a conflict mitigation tool.
Ambiance is a unique fit for the WILDLABS Awards because it addresses key gaps in conservation technology: testing scalable APS applications in real-world contexts, building local manufacturing and technical capacity, and delivering an open-source solution for human-wildlife coexistence. Ultimately, this project aims to transform how communities and wildlife share landscapes, creating a scalable model that could be applied across Africa and beyond.
Project lead: Michael Kowalski, UC Santa Cruz