The importance of collaborative studies between the fields of biology and engineering are increasing. Technological innovation is a primary driver of advances in molecular and cellular biology. Unfortunately, most cutting-edge technologies are difficult to practically apply to biology. In embryogenesis, biological systems show a high degree of spatially controlled differentiation patterns. To understand underpinning mechanisms of spatial differentiation patterns of human mesenchymal stem cells (hMSCs) quantitatively in an in vitro system, both advanced micro-fabrication and image analysis technologies are required. hMSC differentiation patterns induced here by the cultivation of cells in confined space and by exposing them to differentiation induction media. This paper discusses an ensemble of nonadhesive agarose micro cell-culture wells (microwells) confining cells on adhesive substrates and artificial intelligence (AI) to understand why hMSC do not differentiate homogeneously, but in patterns, from the viewpoint of usability of our technological advances in actual high throughput screening experiments.