Characterizing and Modeling AI-Driven Animal Ecology Studies at the Edge

This repo provides instructions for extracting workload information from AI-Driven Animal Ecology (ADAE) studies. We also provide instructions for modelling ADAE studies as time-varying Poisson arrival rates. These simulated workloads can be used to test different scaling techniques (independent and correlated) and validate edge computing systems for ADAE studies in the field.

Figure 1: Workflow of AI-Driven Animal Ecology Studies

Paper Abstract

Platforms that run artificial intelligence (AI) pipelines on edge computing resources are transforming the fields of animal ecology and biodiversity, enabling novel wildlife studies in animals’ natural habitats. With emerging remote sens- ing hardware, e.g., camera traps and drones, and sophisticated AI models in situ, edge computing will be more significant in future AI-driven animal ecology (ADAE) studies. However, the study’s objectives, the species of interest, its behaviors, range, and habitat, and camera placement affect the demand for edge resources at runtime. If edge resources are under-provisioned, studies can miss opportunities to adapt the settings of camera traps and drones to improve the quality and relevance of captured data. This paper presents salient features of ADAE studies that can be used to model latency, throughput objectives, and provision edge resources. Drawing from studies that span over fifty animal species, four geographic locations, and multiple remote sensing methods, we characterized common patterns in ADAE studies, revealing increasingly complex workflows involving various computer vision tasks with strict service level objectives (SLO). ADAE workflow demands will soon exceed individual edge devices’ compute and memory resources, requiring multiple networked edge devices to meet performance demands. We developed a framework to scale traces from prior studies and replay them offline on representative edge platforms, allowing us to capture throughput and latency data across edge configurations. We used the data to calibrate queuing and machine learning models that predict performance on unseen edge configurations, achieving errors as low as 19%

Step 1: Extract arrival rates from real ADAE Studies

We provide links to the raw data linked below. Alternatively, you have use the cleaned data provided in the data directory and skip to Step 2.

Camera trap dataset from LILA BC

Data: Orinoquía Camera Traps
Code: extract_camtrap.py

Drone dataset from KABR

Data: KABR Telemetry
Code: extract_kabr.py

Step 2: Model ADAE workloads

Use change_points.py to extract change points and arrival rates for the time-varying Poisson process.

Visualize and analyze data with the Jupyter notebook provided.

Figure 2: ADAE Studies in the field using networks of camera traps and drones

Acknowledgements

This work was funded by the National Science Foundation (NSF) grant OAC-2112006 (ICICLE AI Institute).