Development and Implementation of a Low-Cost LBL Navigation System for an AUV Aníbal Matos, Nuno Cruz, Alfredo Martins and Fernando Lobo Pereira Faculdade de Engenharia da Universidade do Porto and Instituto de Sistemas e Robótica R. dos Bragas, 4050-123 Porto Portugal {anibal,nacruz,aom,flp}@fe.up.pt Abstract — A reliable navigation system is a key factor for the success of an operational mission with an AUV in a real scenario. In this paper, we address the main issues involved in the implementation of a long baseline (LBL) navigation system for a REMUS AUV. This system replaces both the original hardware and software of the vehicle with a simpler, faster, less expensive and more precise system, based on a Kalman filter. We also discuss the influence of transponder location in the overall performance of the LBL navigation, and present results obtained with this new system in operational missions. I. INTRODUCTION The autonomous operation of an AUV requires an on-board navigation system to compute, in real-time, the position of the vehicle. There are many different systems to determine the position of an AUV [1]. For low cost AUVs an acoustic position system complemented with velocity measurements seems to be the most common solution [1,2,3]. In this paper we describe a new navigation system developed for Isurus, a REMUS class AUV [4]. This is a torpedo shaped vehicle with a diameter of 20 cm and about 1.5 meters long. The propulsion is provided by a rear placed propeller and the vehicle has horizontal and vertical control surfaces to change its position both in the vertical and the horizontal planes. The vehicle is equipped with a set of sensors for navigation and control: a pressure cell to measure the vehicle depth, a digital magnetic compass to determine its heading, two tilt sensors that measure the vehicle roll and pitch angles, a shaft encoder to measure the propeller rotation speed and sensors to determine the angular position of the control surfaces. The vehicle has also an omni-directional acoustic transducer. This transducer is capable of transmitting and receiving acoustic signals in the medium frequency range (20kHz to 30kHz). The new navigation system replaces the original Remus navigation system. The original system could be used to navigate either in LBL or USBL mode and used a DSP to process the acoustic signals received by the vehicle. The new system operates only in LBL mode and the DSP based signal detection was replaced by analog filters tuned to the frequencies of the signals emitted by the transponders. An algorithm that fuses range measurements with dead-reckoning information was also implemented in the on-board computer. This algorithm was implement in the new structure of the on-board software that has been developed for Isurus [5,6], using a real time operating system. This paper is organized as follows. In section 2, we describe the overall structure of the navigation system. In section 3, we discuss the sensitivity of the positioning algorithm as a function of the location of the vehicle with respect to the transponders and of the location of the transponders themselves. In sections 4 and 5, we present the two parts of the developed system: the acoustic signal detection hardware, and the algorithm that merges the dead reckoning information and the range measurements to produce an estimate of the vehicle position. In section 6, we present some results achieved with this system in a real mission performed by the vehicle. II. SYSTEM DESCRIPTION The main function of the developed navigation system is to determine the position of the vehicle in the horizontal plane, as the vertical coordinate of the vehicle is given directly from the pressure cell. To perform such a task, the navigation system uses range measurements to a set of acoustic transponders deployed in the area of operation and information of the vehicle velocity relative to the water. To obtain a range measurement, the vehicle has to transmit an interrogation signal to a given transponder, detect the reply from the transponder and measure the round trip travel time of the acoustic signal. The vehicle velocity is obtained by measuring the propeller rotation speed and the vehicle heading, pitch and roll. Velocity measurements are fused with range measurements by a Kalman filter based algorithm, taking advantage of the characteristics of each type of data. On one hand, the vehicle velocity data is available at a high rate, but its integration leads to a drift in the estimated position. On the other, range measurements, available at a lower rate, can be noisy but do not drift over time. The algorithm updates the estimate of the vehicle position at the same rate the velocity is measured, and