Do we really have to wait until the BionicMobileAssistant is adapted to agriculture to talk about human-robot collaboration, or could we imagine a third way where agriculture itself is rethought for the robot from the start? We ignore the menu by focusing on the costly adaptation of an existing robot to fields designed without it. Why not consider crops or greenhouses where the rows are already spaced for optimal automated harvesting? For example, instead of forcing the robot to pick traditional strawberries, we could develop varieties better suited to mechanical grasping, which would greatly simplify the operational integration.