Greenhouse and orchard fruit harvesting has posed a challenge due to being time-consuming and labor-intensive, the seasonal shortage of skilled laborers, and the aging workforce. The high demand for fresh market produce increases the demand for mechanized and robotic harvesting systems in greenhouses and orchards. Greenhouse automation applications have high demand due to their application success and the efficiency of fruit production. Despite significant development in greenhouse robotic applications, widespread use of state-of-the-art intelligent deep learning-based visual perception applications has yet to be applied at the field level. The trend of orchard architecture, changing and training fruit tree structure, resulted in high yield and easy management practices. Novel orchard architecture helps to replace conventional multi-degree robotic arms with smaller-degree canopy-specific multi-arm robotic systems. This chapter analyzes the existing greenhouse and orchard fruit harvesting systems by examining the appropriateness of novel, smaller-degree multi-arm systems linking with emerging concepts of digital twin and virtual reality application.