Self-regulation (or so-called homeostasis) is a property of all living organisms to maintain an internal stable state through specialized biofeedback mechanisms under varying external and internal conditions. Although these feedback mechanisms in living organisms are complex networks and hard to implement one-to-one in artificial systems, the new approaches in soft robotics may benefit from the concept of self-regulation-especially in the new endeavors of making untethered, autonomous soft robots. In this study, we show a simple system, in which plant robots display heliotropism (sun tracking) and nyctinasty (leaf opening) through artificial self-regulation attained through a bioinspired transpiration mechanism. The feedback involves dehydration/hydration and transpiration events that keep the stem continuously in a metastable position, which maximizes light on plant leaves and the efficiency of light harvesting when solar panels are attached on leaves. We also demonstrate that this artificial feedback can be regulated by doping with light-absorbing chemicals or by changing the geometry of the system, and it can further be expanded to other lightweight systems. Implementing self-regulation into (soft) robots through bioinspired material feedback is beneficial not only for energy efficiency and harvesting but also for achieving embodied intelligence in autonomous soft robots.