and dense built areas. These parameters are unacceptable for manned and autonomous transportation in cities. That is arise the following problems for manned navigation: skipping proper turns and making wrong turns, getting on improper driveway or taking false exit on road junctions, loss of navigation and jumps of position in: urban canyons, financial districts, sky scrapers blocks, narrow and dense built areas; loss of navigation and incorrect positioning in city tunnels, nearby viaducts and overpasses, absence of navigation inside underground, under roof, multilevel parking lots, no seamless outdoor/indoor navigation. All mentioned issues makes customers unsure and unsatisfied.

Autonomous/self-driving From the autonomous/self-driving point of view the goals are: decimeter grade of absolute positioning within a global reference frame, relative localization/location displacement estimation accuracy – 10 cm per 100 sec, speed estimation accuracy – 5 cm/sec, course estimation accuracy – 0,1-0,5 deg. Effective integration of navigation and machine control information sources as GNSS, IMU (gyro, accelerometer), 3D compass, Digital Map, Radar, Lidar, Image-based Positioning and Computer Vision. Solution to the navigation/positioning task mainly within GNSS+IMU+3DMap+Image-based Positioning techs. To allocate the Computer Vision and Radar/Lidar techs mostly for obstacle avoidance, objects recognition and machine control functions.

Urbanavtic technology Urbanavtic approach to urban mobility and autonomous/self-driving technology solution concludes in deep engineering synthetic navigation design consist of: tight couple integration of GNSS with internal sensors (IMU) using raw data of satellite receiver, map based corrections (road graph and turns matching), image-based positioning (reference points detection and images recognition; no permanent scene analysis typical for obstacle avoidance, no high power and computation consumption). The software only product using just a platform sensors running on smartphones and in-dash auto systems.

Auto insurance Auto insurance problem: it is a must to have precise motion trajectories of manned and autonomous vehicles’ in absolute and relative coordinates in case of collision happened. Traditional and unmanned vehicles collide today and will collide in future for various reasons. Perfectly reconstructed trajectories based on robust motion data opens the opportunity for automatic arbitration and resolution of road traffic incidents and accidents. Achievable centimeter accuracy of relative and decimeter accuracy of absolute positioning is the core technology for automatic collisions arbitration. High labor costs of “manual” non-automatic settlement of traffic accidents and insurance cases nowadays. Bottom line: no clear picture of road traffic accident, no data to judge who is responsible for crash, no data for crash analysis.

Urbanavtic technology Urbanavtic’s auto insurance technology solution is founded on: precise relative positioning (estimation of the location displacement) based on phase measurements of GNSS and transferring to absolute coordinates using map matching methods and image-based positioning. Here we use applied measurements: GNSS raw data, primarily the accumulated phases and MEMS IMU sensors, image frames from camera. Dynamic adaptation of data processing based on a constant analysis of consistency of measurements from the phase tracking channels. Exact reconstruction of vehicles’ trajectories based on entire set of kinematic parameters including accelerations, velocities and angular velocities. Hereby we have opened the opportunity for automatic arbitration of road traffic accidents, cut costs of insurance companies, save time and money for drivers.

Demonstration of results The video of the technology real time tests on Android smartphone will be demonstrated in the frame of presentation. Interested customers will be welcome for field tests.