Agriculture is one of the most important economic sectors and essential for our daily life. Especially in relation to selectively harvested crops, agriculture is majorly affected by increasing wage demands and workforce shortages. While other industries tackle this issue with increasing automation, the task of harvesting fruit and vegetables is still largely fulfilled manually. Following an overview of selective harvesting robot architecture, this paper addresses the question of which crops are ideally suited for AI-assisted selective harvesting. For this purpose, the state of development of robots for specific crops is evaluated based on performance measurements, and crop characteristics are determined through interviews with domain experts. We found that cauliflower and broccoli have limited suitability, sweet peppers, apple and cucumber have good suitability, and kiwi, strawberry and tomato are most suitable.