Rajmohan Madhavan

Iterative registration of 3D LADAR data for autonomous navigation

This paper describes an iterative algorithm for registration of 3D LADAR data. The proposed approach is iconic in nature with suitable modifications to deal with false/spurious matches, occlusions and outliers. Experimental results using data obtained from field trials on an experimental unmanned vehicle (XUV) are presented to demonstrate the efficacy of the approach. The paper also details ongoing research efforts to determine the feasibility of employing the algorithm for real-time autonomous navigation.

Applications of a 3D Range Camera Towards Healthcare Mobility Aids

The National Institute of Standards and Technology (NIST) has recently studied a new 3D range camera for use on mobile robots. These robots have potential applications in manufacturing, healthcare and perhaps several other service related areas beyond the scope of this paper. In manufacturing, the 3D range camera shows promise for standard size obstacle detection possibly augmenting existing safety systems on automated guided vehicles. We studied the use of this new 3D range imaging camera for advancing safety standards for automated guided vehicles. In healthcare, these cameras show promise for guiding the blind and assisting the disabled who are wheelchair dependent. Further development beyond standards efforts allowed NIST to combine the 3D camera with stereo audio feedback to help the blind or visually impaired to stereophonically hear where a clear path is from room to room as objects were detected with the camera. This paper describes the 3D range camera and the control algorithm that combines the camera with stereo audio to help guide people around objects, including the detection of low hanging objects typically undetected by a white cane

Temporal range registration for unmanned ground and aerial vehicles

An iterative temporal registration algorithm is presented in this paper for registering 3D range images obtained from unmanned ground and aerial vehicles traversing unstructured environments. We are primarily motivated by the development of 3D registration algorithms to overcome both the unavailability and unreliability of Global Positioning System (GPS) within required accuracy bounds for unmanned ground vehicle (UGV) navigation. After suitable modifications to the well-known iterative closest point (ICP) algorithm, the modified algorithm is shown to be robust to outliers and false matches during the registration of successive range images obtained from a scanning LADAR rangefinder on the UGV. Towards registering LADAR images from the UGV with those from an unmanned aerial vehicle (UAV) that flies over the terrain being traversed, we then propose a hybrid registration approach. In this approach to air to ground registration to estimate and update the position of the UGV, we register range data from two LADARs by combining a feature-based method with the aforementioned modified ICP algorithm. Registration of range data guarantees an estimate of the vehicles position even when only one of the vehicles has GPS information. Temporal range registration enables position information to be continually maintained even when both vehicles can no longer maintain GPS contact. We present results of the registration algorithm in rugged terrain and urban environments using real field data acquired from two different LADARs on the UGV.

Towards AGV safety and navigation advancement obstacle detection using a TOF range camera

The performance evaluation of an obstacle detection and segmentation algorithm for automated guided vehicle (AGV) navigation using a 3D real-time range camera is the subject of this paper. Our approach has teen tested successfully on British safety standard recommended object sizes and materials placed on the vehicle path. The segmented (mapped) obstacles are then verified using absolute measurements obtained using a relatively accurate 2D scanning laser rangefinder. Sensor mounting and sensor modulation issues are also described through representative data sets

Moving Object Prediction for Off-road Autonomous Navigation

The realization of on- and off-road autonomous navigation of Unmanned Ground Vehicles (UGVs) requires real-time motion planning in the presence of dynamic objects with unknown trajectories. To successfully plan paths and to navigate in an unstructured environment, the UGVs should have the difficult and computationally intensive competency to predict the future locations of moving objects that could interfere with its path. This paper details the development of a combined probabilistic object classification and estimation theoretic framework to predict the future location of moving objects, along with an associated uncertainty measure. The development of a moving object testbed that facilitates the testing of different representations and prediction algorithms in an implementation-independent platform is also outlined.

Autonomous Bayesian Search and Tracking, and its Experimental Validation

We present a technique that uniformly controls a team of autonomous sensor platforms charged with the dual task of searching for and then tracking a moving target within a recursive Bayesian estimation framework. The proposed technique defines the target detectable region, and uniformly formulates observation likelihoods with detection and no-detection events. The unified likelihood function allows the proposed technique to update and maintain the target belief, regardless of the target detectability. For unified search and tracking (SAT), the proposed technique further predicts the belief in a finite-time horizon, and decides control actions by maximizing a unified objective function consisting of local and global measures derived from the predicted belief. Using the objective function, the proposed technique can smoothly change its control actions even during transitions between SAT. The numerical results first show successful SAT by the proposed technique in tests using a sensor platform with different detectability and comparison with conventional searching techniques under different prior knowledge, and then identifies the superiorities of the proposed technique in SAT. The experimental results finally validate the applicability and extendability of the proposed technique via coordinated SAT in a field experiment.

A hierarchical, multi-resolutional moving object prediction approach for autonomous on-road driving

In this paper, we present a hierarchical multi-resolutional approach for moving object prediction via estimation-theoretic and situation-based probabilistic techniques. The results of the prediction are made available to a planner to allow it to make accurate plans in the presence of a dynamic environment. We have applied this approach to an on-road driving control hierarchy being developed as part of the DARPA Mobile Autonomous Robotic Systems (MARS) effort. Experimental results are shown in two simulation environments.

Stable Navigation Solutions for Robots in Complex Environments

During the initial phase of a disaster response it is essential for responders to quickly and safely assess the overall situation. The use of rescue robots that can autonomously navigate and map these environments can help responders realize this goal while minimizing danger to them. In order for rescue robots to be of service to the responders, they must be able to sense the environment, create an internal representation that identifies victims and hazards to responders, and provide an estimate of where they are and where they have been. Methods for developing a stable navigation solution are based on sensors that can be broadly classified into two approaches, absolute (exteroception) and relative (proprioception). Commonly, two or more of these approaches are combined to develop a stable navigation solution that is insensitive to and robust in the presence of the errors that plague partial solutions by taking into account errors in the vehicles pose, thus bounding the uncertainty in the navigation solution. Since the capabilities and limitations of these approaches vary, it is essential for developers of robotic systems to understand the performance characteristics of methodologies employed to produce a stable navigation solution. This paper will provide quantitative analysis of two proprioceptive approaches, namely encoder-based odometry and inertial navigation system, and an exteroceptive approach namely visual odometry that uses scan matching techniques.

Advancing manufacturing research through competitions

Competitions provide a technique for building interest and collaboration in targeted research areas. This paper will present a new competition that aims to increase collaboration amongst Universities, automation end-users, and automation manufacturers through a virtual competition. The virtual nature of the competition allows for reduced infrastructure requirements while maintaining realism in both the robotic equipment deployed and the scenarios. Details of the virtual environment as well as the competitions objectives, rules, and scoring metrics will be presented.

Information-based intelligent unmanned ground vehicle navigation

Sensor-centric navigation of unmanned ground vehicles (UGVs) operating in rugged and expansive terrains requires the competency to evaluate the utility of sensor information such that it results in intelligent behavior of the vehicles. In this paper, we propose an entropic information metric for the above purpose where entropy is used to quantify the probabilistic uncertainty in sensor measurements. We present results using data obtained from field trials on an unmanned vehicle to substantiate the utility of the proposed metric. We also show how low and high level tasks can be predicated upon this metric in potential application areas related to autonomous vehicle navigation.