There has been much research on social robots that provide information while conversing with people in public facilities and homes. I have developed and demonstrated various robot systems, including a guide robot that explains exhibits at a science museum, a care robot that chats with older adults at a nursing home, and entertainment robots that discuss human nature with people at an event. In terms of time and cost, such demonstrations in real-world environments prefer old-fashioned but robust methods (e.g., dialogue management based on manual scenarios) rather than sophisticated methods (e.g., state-of-the-art machine learning technology). As a result of my involvement in those projects, I have acquired the know-how to create scenarios for human-robot interaction. That know-how may be insignificant in this AI era, but it can be helpful to implement interactive robots working in the real world quickly. In this talk, I will talk about this know-how.