Subterranean environments, including tunnels, sewers, and water pipes, present a complex setting with multiple environmental and structural hazards as their conditions degrade and change over time which are too risky situations for humans. Moreover, inspection and maintenance of these environments require road excavations which causes road closures, in addition, to being costly, noisy, dirty, and inconvenient. To cope with these issues, infrastructure robots can be utilised to do tasks such as checking the degradation condition and leak localisation and doing reparation. However, these environments are characterized by darkness, moisture, temperature, and water conditions, isolation and restriction in space, and curvatures. These conditions present significant challenges in the robot structure design, control, localisation, navigation, and communication. Most current research on subterranean robots focuses on using a single robot to do inspection tasks. For data acquisition and supervisory control, a communication link between the base station and the robot is mandatory. This link must be a wireless channel due to the subterranean environment geometry. However, the above-mentioned characteristics make communication unreliable where Non-Line-of-Sight (NLOS) occurs. This project presents a multi-robot system to save the time and money that goes into the maintenance of these environments considering the above-mentioned hazards and challenges.