We present a novel way to learn sampling distributions for sampling-based motion planners by making use of expert data. We learn an estimate (in a non-parametric setting) of sample densities around semantic regions of interest, and incorporate these learned distributions into a sampling-based planner to produce natural plans. Our motivation is that certain aspects of the workspace have a local influence on planning strategies, which is dependent both on where, and what, they are. In the event that learning the density estimate of the training data is impractical in the original feature space, we utilize a non-linear dimensionality-reduction technique and perform density estimation on a lower-dimensional embedding. Samples are then lifted from this embedded density into the original feature space, producing samples that still well approximate the original distribution.

A goal of this work is to learn how various features in the environment influence the behavior of experts – for example, how pedestrian crossings, traffic signals and so on affect drivers. We show that learning sampling distributions from expert trajectory data around these semantic regions leads to more natural paths that are measurably closer to those of an expert. We demonstrate the feasibility of the technique in various scenarios for a virtual car-like robotic vehicle and a simple manipulator, contrasting the differences in planned trajectories of the semantically-biased distributions with conventional techniques.


  • [PDF] I. Baldwin and P. Newman, “Non-parametric Learning for Natural Plan Generation,” in International Conference on Intelligent Robots and Systems, Taipei, Taiwan, 2010.
    author = {Ian Baldwin and Paul Newman},
    title = {Non-parametric Learning for Natural Plan Generation},
    booktitle = {International Conference on Intelligent Robots and Systems},
    year = {2010},
    address = {Taipei, Taiwan},
    month = {October},
    note = {10},
    pdf = {http://www.robots.ox.ac.uk/~mobile/Papers/BaldwinIROS2010.pdf},
    keywords = {Learning to Plan},

 The figures below validate the process on a manipulator-planning problem for a standard platform, the Puma 560. The goal is to move the end-effector from the initial position (shown in (a)) to the goal position (marked in green) over the top of the obstacle (shown in grey).  An expert solution is whose in (b).


Screen Shot 2014-11-17 at 11.37.25

Solutions to the manipulator task of both the semantically-biased planner (blue ) and a standard RRT solution (green) are shown in (c). As can be seen, the semantic-planner produces paths more similar to the expert – the semantic planner has learned to plan “over” the box. Screen Shot 2014-11-17 at 11.37.31