Accurate Position Estimation of a Drifting Wireless LAN Communication Device in a 200mm-Diameter Small Sewer Pipe

Abstract

Sewer pipe deterioration is a severe problem in many countries, and sewer pipe inspections are essential for maintaining sewer systems. Various sewer inspection methods, such as a visual check, boat-type video cameras, and remote robots with or without a wired connection, exist. However, they each suffer from problems, such as high labor costs, long waiting times when using boat-type video cameras, and the purchasing cost of robots. To achieve low-cost, safe, and near real-time inspections of sewer pipes, we have proposed a sewer inspection method based on a drifting wireless camera and sensor nodes. Drifting nodes capture a video of the inside of pipe while drifting on a current in the sewer and access points (AP) installed under manholes receive video data from drifting nodes and send data to an online server on the Internet. For the video inspection using our proposed system, the information of the position of camera/sensor nodes while they drifting is essential because video data need to link to the location information to help to detect damaged points of pipes. However, the camera/sensor nodes cannot utilize the GNSS because signals from GNSS satellite do not reach to the space in underground sewer pipes. In this paper, to investigate the usefulness of the receive signal strength indicator (RSSI) observed by drifting camera/sensor nodes for the position estimation in 200mm-diameter pipes, we performed experiments to measure the RSSI of a wireless communication device drifting on water in an experimental underground pipe. The measurement results show that RSSI monotonically increases when a drifting device approaches the AP and monotonically decreases when it leaves the AP, and the distribution of measurement value suggests that possibility of position estimation with an error of lm order.