Animal farming has transitioned from small-scale operations to large commercial ventures, raising concerns about animal welfare alongside productivity and profitability. Artificial intelligence technologies offer significant potential for enhancing welfare through improved monitoring. However, practical solutions for optimizing farm management decisions are still limited. This paper tackles the challenge of optimizing camera placement in chicken farms to achieve maximum coverage for effective monitoring. We employ the Covariance Matrix Adaptation Evolution Strategy (CMA-ES) and two Quality Diversity (QD) algorithms, MAP-Elites (ME) and CMA-ME, using two behaviour descriptors to identify optimal camera positions. Our findings show that algorithm-derived camera placements outperform human-designed configurations. Importantly, the QD algorithms offer a diverse set of high-quality solutions that can be selected without extra computation, in case the CMA-ES solution does not meet unforeseen constraints. Using our modelling, we have optimized the camera placement in a commercial farm in Cyprus offering maximum coverage surveillance.