In the last years considerable emphasis has been placed on safety at sea. There is the maritime security and surveillance system whose main aim is to execute tasks in the interests of maritime safety and to react in case of emergency. They are monitored by networks of radar stations. On such areas we obtain a lot of navigation data which could be used to improve ship’s parameters (position), using know in geodesy modern M-estimation methods. Simultaneous acquisition of navigational information from many independent radar stations will render it possible to obtain a more accurate ship position in marine traffic surveillance systems in relation to the calculated position. A position expected in an adjustment calculus is received from a watch officer. It is burdened with a fallacy of navigation systems and the quality of marking ship’s route on a map. In the case of navigational-parameter measurements used for depicting ship position, one can obtain incorrect results due to a disturbance in the measurement process. In extreme cases, such erroneous data could significantly differ from the anticipated results. Deviating observations could significantly influence the values of measurement results. In order to eliminate the determination of erroneous measurements, one could use resistant estimation methods with suitably selected attenuation functions. The accuracy of a determined position will not be better than the capabilities of the device used. Adjustment gives the possibility of eliminating or minimizing human errors as well as the errors in the indications of navigational devices. This paper presents the latest robust estimation methods using Danish attenuation function for adjustment of navigational observation, using radar observation.