Impact of Image Resolution on Drone Surveillance System Availability: A Stochastic Petri Net Approach

• Ivson Borges UFPE
• Luan Lins UFPE
• Gustavo Callou UFRPE
• Paulo Maciel UFPE

Resumo

Ensuring the availability and energy autonomy of drone-based surveillance systems is critical as these technologies become increasingly vital for security, disaster response, and public safety operations. The effectiveness of such systems depends not only on the reliability of their components but also on the trade-off between energy consumption and the performance of onboard artificial intelligence (AI) algorithms, which are highly sensitive to image resolution. This paper presents a stochastic modeling approach to analyze how different image resolution settings impact the operational availability of UAV-based surveillance systems. Stochastic Petri Net (SPN) models are proposed to evaluate the effects of drone and battery redundancy strategies on system availability. The model incorporates mission-critical parameters such as drone failure and repair rates, battery discharge and recharge cycles, and energy consumption as a function of image resolution. We examine the metrics under evaluation assuming Full HD, 2K, and 4K image resolutions. The results show that battery redundancy significantly improves UAV system availability, while adding a backup drone yields marginal gains, offering valuable design insights to balance energy efficiency, resolution, and operational reliability.