Inside Unmanned Systems provides actionable business intelligence to decision-makers and influencers operating within the global UAS community. Features include analysis of key technologies, policy/regulatory developments and new product design.
Issue link: https://insideunmanned.epubxp.com/i/668560
62 unmanned systems inside April/May 2016 GROUND AUTOMOTIVE grade and does not require the system to be re-designed. A class for this new measurement type is developed and uploaded to the RIFE library. After that, aiding sensors that corre- spond to this class can be used in a plug and play manner as described above. The RIFE design also supports sensor data with latencies and timing inconsistencies. A generic backward-propagation method is developed so as to incorporate asynchronous and out-of-sequence measurements. The system operates on a DRN measurement cycle. If any aiding measurements become available since last DRN update, they are incorporated into RIFE aiding observables and applied to update the filter states. DRN states are propagated backwards to the aiding validity time where aiding observables are formulated. The state propagation is followed by the backward propagation of error states and covariance matri- ces similar to the filter prediction update. Application Case Study As a specific case study, the plug and play navi- gation technology is applied for automotive vehicle-to-vehicle (V2V) safety systems, whose adoption by the automotive mass market is ex- pected to start within next three to five years. Figure 7 shows a generalized architecture of a V2V safety system. An automotive navigation module (ANM) performs real-time estimation of vehicle posi- tion and velocity states. Inter-vehicle exchange of navigation data is enabled by a transponder, which uses dedicated short-range communica- tion (DSRC) radios to support the data exchange between neighboring vehicles (generally, within a 200-250 meter range). A controller compares the vehicle's own trajectory with trajectories of surrounding vehicles in order to predict poten- tial collisions. Collision prediction results are used to generate audio, visual, and/or haptic warning signals to the driver. Operating in an advisory mode, the V2V safe- ty system does not perform automatic collision avoidance per se. Rather the final decision is Figure 7: Generalized architecture of the V2V safety system Figure 8: Examples of accidents prevented by the V2V Figure 9: Ground vehicle test setup for demonstration of plug-and-play navigation capabilities Figure 10: Annotated sensor board Left turn on at unregulated intersection when the visibility of oncoming traf f c is (partially) obstructed, which leads to a possibility of side collision Passing on a two-lane road with a limited visibility of the oncoming traf f c, which leads to a possibility of head-on collision GPS antenna Video cameras a. b. b. a.