Continuity risk of feature extraction for laser-based navigation

In this paper, a new continuity risk evaluation method is developed, simulated, and tested for laser-based navigation algorithms using feature extraction (FE) and data association (DA). A major risk for safety in FE and DA is caused by incorrect association, which happens when attributing a feature extracted from laser data to the wrong landmark in a pre-established map. In prior work, we designed an innovation-based DA process to evaluate the integrity risk caused by incorrect associations while considering all potential measurement permutations. In this paper, permutations are used again at the FE step to determine the minimum normalized separation between extracted features. Features that are poorly separated are easily found, but are likely to be incorrectly associated. If the minimum separation is smaller than expected, then features are not extracted, which causes loss of navigation continuity. This paper provides an analytical upper-bound on the continuity risk caused by nominal laser measurement errors, and an integrity risk bound guaranteeing a predefined level of continuity. These safety risk bounds are analyzed and tested in example scenarios showing that the lower the continuity risk requirement is, the higher the integrity risk due to incorrect associations becomes.