Mobile robots can access regions and collect data in structural locations not easily reached by humans. This includes confined spaces, such as inside walls, and underground pipes; and remote spaces, such as the underside of bridge decks. Robot access provides the opportunity to sense in these difficult to access spaces with robot mounted sensors, i.e. cameras and lidars, and with the robot placing and servicing standalone sensors. Teams of robots, sensors and AR-equipped humans have the potential to provide rapid and more comprehensive structural assessments. This paper presents results of studies using small robots to explore and collect structural condition data from remote and confined spaces including in walls, culverts, and bridge deck undersides. The presentation also covers system and network architecture, methods for automating data processing with localized and edge-based processors, the use of augmented reality (AR) human interfaces and discusses key technical challenges and possible solutions.