Kristoffer M. Frey

Kristoffer M. Frey, Ted J. Steiner, Jonathan P. How

Demand for high-performance, robust, and safe autonomous systems has grown substantially in recent years. These objectives motivate the desire for efficient safety-theoretic reasoning that can be embedded in core decision-making tasks such as motion planning, particularly in constrained environments. On one hand, Monte-Carlo (MC) and other sampling-based techniques provide accurate collision probability estimates for a wide variety of motion models but are cumbersome in the context of continuous optimization. On the other, "direct" approximations aim to compute (or upper-bound) the failure probability as a smooth function of the decision variables, and thus are convenient for optimization. However, existing direct approaches fundamentally assume discrete-time dynamics and can perform unpredictably when applied to continuous-time systems ubiquitous in the real world, often manifesting as severe conservatism. State-of-the-art attempts to address this within a conventional discrete-time framework require additional Gaussianity approximations that ultimately produce inconsistency of their own. In this paper we take a fundamentally different approach, deriving a risk approximation framework directly in continuous time and producing a lightweight estimate that actually converges as the underlying discretization is refined. Our approximation is shown to significantly outperform state-of-the-art techniques in replicating the MC estimate while maintaining the functional and computational benefits of a direct method. This enables robust, risk-aware, continuous motion-planning for a broad class of nonlinear and/or partially-observable systems.

Kristoffer M. Frey, Ted J. Steiner, Jonathan P. How

As autonomous systems increasingly rely on onboard sensing for localization and perception, the parallel tasks of motion planning and state estimation become more strongly coupled. This coupling is well-captured by augmenting the planning objective with a posterior-covariance penalty -- however, prediction of the estimator covariance is challenging when the observation model depends on unknown landmarks, as is the case in Simultaneous Localization and Mapping (SLAM). This paper addresses these challenges in the case of landmark- and SLAM-based estimators, enabling efficient prediction (and ultimately minimization) of this performance metric. First, we provide an interval-based filtering approximation of the SLAM inference process which allows for recursive propagation of the ego-covariance while avoiding the quadratic complexity of explicitly tracking landmark uncertainty. Secondly, we introduce a Lie-derivative measurement bundling scheme that simplifies the recursive "bundled" update, representing significant computational savings for high-rate sensors such as cameras. Finally, we identify a large class of measurement models (which includes orthographic camera projection) for which the contributions from each landmark can be directly combined, making evaluation of the information gained at each timestep (nearly) independent of the number of landmarks. This also enables the generalization from finite sets of landmarks $\{\ell^{(n)} \}$ to distributions, foregoing the need for fully-specified linearization points at planning time and allowing for new landmarks to be anticipated. Taken together, these contributions allow SLAM performance to be accurately and efficiently predicted, paving the way for online, observability-aware trajectory optimization in unknown space.

Kristoffer M. Frey, Ted J. Steiner, Jonathan P. How

Data association in SLAM is fundamentally challenging, and handling ambiguity well is crucial to achieve robust operation in real-world environments. When ambiguous measurements arise, conservatism often mandates that the measurement is discarded or a new landmark is initialized rather than risking an incorrect association. To address the inevitable `duplicate' landmarks that arise, we present an efficient map-merging framework to detect duplicate constellations of landmarks, providing a high-confidence loop-closure mechanism well-suited for object-level SLAM. This approach uses an incrementally-computable approximation of landmark uncertainty that only depends on local information in the SLAM graph, avoiding expensive recovery of the full system covariance matrix. This enables a search based on geometric consistency (GC) (rather than full joint compatibility (JC)) that inexpensively reduces the search space to a handful of `best' hypotheses. Furthermore, we reformulate the commonly-used interpretation tree to allow for more efficient integration of clique-based pairwise compatibility, accelerating the branch-and-bound max-cardinality search. Our method is demonstrated to match the performance of full JC methods at significantly-reduced computational cost, facilitating robust object-based loop-closure over large SLAM problems.

Kristoffer M. Frey, Ted J. Steiner, Jonathan P. How

Sparsity has been widely recognized as crucial for efficient optimization in graph-based SLAM. Because the sparsity and structure of the SLAM graph reflect the set of incorporated measurements, many methods for sparsification have been proposed in hopes of reducing computation. These methods often focus narrowly on reducing edge count without regard for structure at a global level. Such structurally-naive techniques can fail to produce significant computational savings, even after aggressive pruning. In contrast, simple heuristics such as measurement decimation and keyframing are known empirically to produce significant computation reductions. To demonstrate why, we propose a quantitative metric called elimination complexity (EC) that bridges the existing analytic gap between graph structure and computation. EC quantifies the complexity of the primary computational bottleneck: the factorization step of a Gauss-Newton iteration. Using this metric, we show rigorously that decimation and keyframing impose favorable global structures and therefore achieve computation reductions on the order of $r^2/9$ and $r^3$, respectively, where $r$ is the pruning rate. We additionally present numerical results showing EC provides a good approximation of computation in both batch and incremental (iSAM2) optimization and demonstrate that pruning methods promoting globally-efficient structure outperform those that do not.