Electronics skin (e-skin) for robots nowadays employs a variety of powerful data processing, complicated electronics, smart materials and structures and transduction techniques. It helps robots in avoiding environmental hazards by providing multiple modalities of sensations similar to the human skin. But, the successful implementation of e-skin in robotics has so far been slow to develop despite being extremely valuable in itself. This article reviews the current e-skin designs and open research questions in the field, focusing on the efficient use of e-skin in robotic systems. The use of e-skin in robotics has a few unique characteristics. For example, unlike other robot senses, the e-skin must be distributed throughout the robot body, much like human skin does. This calls for a distributed and decentralized control and simultaneous processing of several tactile encounters. Because of these constraints, making tactile sensor modalities a reality is difficult. Dealing with concerns including sensor location, electronic/mechanical hardware, signal acquisition techniques, automated calibration procedures and algorithms to analyze and interpret real-time sensor data are necessary to meet these obstacles. Since the performance of a system often depends on how its parts are integrated, we explore this topic from a system viewpoint. Researchers developing robots with touch sensors and system designers are expected to benefit from the survey.