Abstract:Understanding the dynamics of animal social systems requires studying variation in contact and interaction, which is influenced by environmental conditions, resource availability, and predation risk, among other factors. Traditional (direct) observational methods have limitations, but advancements in sensor technologies and data analytics provide unprecedented opportunities to study these complex systems in naturalistic environments. Proximity logging and tracking devices, capturing movement, temperature, and social interactions, offer non-invasive means to quantify behavior and develop empirical models of animal social networks. However, challenges remain in integrating different data types, incorporating more sensor modalities, and addressing logistical constraints. To address these gaps, we developed a wireless wearable sensor system with novel features (called “Juxta”), including modular battery packs, memory management for combining data types, reconfigurable deployment modes, and a smartphone app for data collection. We present data from a pilot study on prairie voles ( ), which is a small mammal species that exhibits relatively complex social behavior. We demonstrate the potential for Juxta to increase our understanding of the social networks and behavior of free-living animals. Additionally, we propose a framework to guide future research in merging temporal, spatial, and event-driven data. By leveraging wireless technology, battery efficiency, and smart sensing modalities, our wearable ecosystem offers a scalable solution for real-time, high-resolution data capture and analysis in animal social network studies, opening new avenues for exploring complex social dynamics across species and environments.

What problem does this paper attempt to address?

The paper aims to address several key issues in the study of animal social networks. Specifically: 1. **Data Integration Challenges**: How to unify time-series data (such as accelerometer data) with event-driven data (such as proximity detection), as these data streams have different granularities and need to be integrated. 2. **Multimodal Sensor Fusion**: How to incorporate more sensor modalities in the study to provide more contextual information about social interactions. For example, combining temperature and motion data can estimate sleep-wake cycles, thereby better understanding the impact of environmental features and internal states on animal social behavior. 3. **Practicality and Portability of Devices**: Current devices face practical issues in field applications, such as battery life, weight, form factor, durability, and the convenience of deployment and retrieval. To address these issues, the authors developed a wireless wearable sensor system called "Juxta." This system has the following features: - Modular battery pack - Memory management combining different types of data - Reconfigurable deployment modes - Smartphone application for data collection By using this new sensor system, researchers were able to capture social interaction data of Prairie Voles in natural environments and demonstrated Juxta's potential in enhancing our understanding of free-living animal social networks and behavior. Additionally, the authors proposed a framework to guide future integration of time, space, and event-driven data, thereby advancing the field further.

A Wireless Wearable Ecosystem for Social Network Analysis in Free-living Animals