Benjamin Grocholsky’s research while affiliated with Carnegie Mellon University and other places

Rotorcraft operating in unprepared locations must be capable of safe flight in the presence of unmapped obstacles and absence of GPS, and must be able to quickly and reliably assess a potential landing zone's suitability for landing. This paper presents technologies that address these needs and the result of their application to an actual unmanned helicopter prototype.

Landing rotorcraft on a ship deck is a difficult and dangerous task. The US Navy is interested in expanding landing capabilities in degraded visual environments, with impaired or no GPS signal, and in autonomous operations, while at the same time reducing the cost of guidance infrastructure on the ship deck. This paper describes how a suite of multi-modal sensors can provide relative pose estimate from an aircraft to a ship deck in a wide range of conditions. The sensor suite enables robust performance while requiring minimal deck side infrastructure. We describe a three-phase trajectory planner that allow for safe, autonomous landing on a ship deck based on the relative pose estimate from the sensor suite and knowledge of the aircraft dynamics. At the aircraft approaches the ship, the trajectory planner uses a ship deck motion model to time the landing for minimal touchdown impact.

The Navy and Marine Corps will increasingly need to operate unmanned air vehicles from ships at sea. Fused multi-sensor systems are desirable to ensure these operations are highly reliable under the most demanding at-sea conditions, particularly in degraded visual environments. The US Navy Sea-Based Automated Launch & Recovery System (SALRS) program aims at enabling automated/semi-automated launch and recovery of sea-based, manned and unmanned, fixed-and rotary-wing naval aircraft, and to utilize automated or pilot-augmented flight mechanics for carefree shipboard operations. This paper describes the goals and current results of SALRS Phase 1, which aims at understanding the capabilities and limitations of various sensor types through sensor characterization, modeling, and simulation, and assessing how the sensor models can be used for aircraft navigation to provide sufficient accuracy, integrity, continuity, and availability across all anticipated maritime conditions.

Coordination of unmanned aerial and ground vehicles (UAVs and UGVs) is immensely useful in a variety of surveillance and rescue applications, as the vehicles’ complementary strengths provide operating teams with enhanced mission capabilities. While many of today’s systems require independent control stations, necessitating arduous manual coordination between mul- tiple operators, this paper presents a multi-robot collaboration system, jointly developed by iRobot Corporation and Carnegie Mellon University, which features a unified interface for controlling multiple unmanned ve- hicles. Semi-autonomous subtasks can be directly executed through this interface, including: single-click automatic visual target tracking, way- point sequences, area search, and geo-location of tracked points of inter- est. Demonstrations of these capabilities on widely-deployed commercial unmanned vehicles are presented, including the use of UAVs as a commu- nication relay for multi-kilometer, non-line-of-sight operation of UGVs.

Comprehensive Automation for Specialty Crops is a project focused on the needs of the specialty crops sector, with a focus on apples and nursery trees. The project’s main thrusts are the integration of robotics technology and plant science; understanding and overcoming socio-economic barriers to technology adoption; and making the results available to growers and stakeholders through a nationwide outreach program. In this article, we present the results obtained and lessons learned in the first year of the project with a reconfigurable mobility infrastructure for autonomous farm driving. We then present sensor systems developed to enable three real-world agricultural applications—insect monitoring, crop load scouting, and caliper measurement—and discuss how they can be deployed autonomously to yield increased production efficiency and reduced labor costs. KeywordsSpecialty crops-Reconfigurable mobility-Crop intelligence-Insect monitoring-Crop load estimation-Caliper measurement

This article compares several parameterizations and motion models for improving the estimation of the nonlinear uncertainty distribution produced by robot motion. In previous work, we have shown that the use of a modified polar parameterization provides a way to represent nonlinear measurements distributions in the Cartesian space as linear distributions in polar space. Following the same reasoning, we present a motion model extension that utilizes the same polar parameterization to achieve improved modeling of mobile robot motion in between measurements, gaining robustness with no additional overhead. We present both simulated and experimental results to validate the effectiveness of our approach.

Coordinated operations between unmanned air and ground assets allow leveraging of multi-domain sensing and increase opportunities for improving line of sight communications. While numerous military missions would benefit from coordinated UAV-UGV operations, foundational capabilities that integrate stove-piped tactical systems and share available sensor data are required and not yet available. iRobot, AeroVironment, and Carnegie Mellon University are working together, partially SBIR-funded through ARDEC's small unit network lethality initiative, to develop collaborative capabilities for surveillance, targeting, and improved communications based on PackBot UGV and Raven UAV platforms. We integrate newly available technologies into computational, vision, and communications payloads and develop sensing algorithms to support vision-based target tracking. We first simulated and then applied onto real tactical platforms an implementation of Decentralized Data Fusion, a novel technique for fusing track estimates from PackBot and Raven platforms for a moving target in an open environment. In addition, system integration with AeroVironment's Digital Data Link onto both air and ground platforms has extended our capabilities in communications range to operate the PackBot as well as in increased video and data throughput. The system is brought together through a unified Operator Control Unit (OCU) for the PackBot and Raven that provides simultaneous waypoint navigation and traditional teleoperation. We also present several recent capability accomplishments toward PackBot-Raven coordinated operations, including single OCU display design and operation, early target track results, and Digital Data Link integration efforts, as well as our near-term capability goals.

A key problem in the deployment of sensor networks is that of determining the location of each sensor such that subsequent data gathered can be registered. We would also like the network to provide localization for mobile entities, allowing them to navigate and explore the environment. In this paper, we present a robust decentralized algorithm for mapping the nodes in a sparsely connected sensor network using range- only measurements and odometry from a mobile robot. Our approach utilizes an extended Kalman filter (EKF) in polar space allowing us to model the nonlinearities within the range-only measurements using Gaussian distributions. We also extend this unimodal centralized EKF to a multi-modal decentralized framework enabling us to accurately model the ambiguities in range-based position estimation. Each node within the network estimates its position along with its neighbor's position and uses a message-passing algorithm to propagate its belief to its neighbors. Thus, the global network localization problem is solved in pieces, by each node independently estimating its local network, greatly reducing the computation done by each node. We demonstrate the effectiveness of our approach using simulated and real-world experiments with little to no prior information about the node locations.

Fielded military unmanned systems are currently extending the reach of the U.S. forces in surveillance and reconnaissance missions. Providing long-range eyes on enemy operations, unmanned aerial vehicles (UAVs), such as the AeroVironment Raven, have proven themselves indispensable without risking soldiers' lives. Meanwhile, unmanned ground vehicles (UGVs), such as the iRobot PackBot, are quickly joining ranks in Explosive Ordnance Disposal (EOD) missions to identify and dispose of ordnance or to clear roads and buildings. UAV-UGV collaboration and the benefit of force multiplication is increasingly more tangible. iRobot Corporation and CMU Robotics Institute are developing the capability to simultaneously control the Raven small UAV (SUAV) and PackBot UGV from a single operator control unit (OCU) via waypoint navigation. Techniques to support autonomous collaboration for pursuing and tracking a dismounted soldier will be developed and integrated on a Raven-PackBot team. The Raven will survey an area and geolocate an operator-selected target. The Raven will share this target location with the PackBot and together they will collaboratively pursue the target intelligently to maintain track on the target. We will accomplish this goal by implementing a decentralized control and data fusion software architecture. The PackBot will be equipped with on-board waypoint navigation algorithms, a Navigator Payload containing a stereo-vision system, GPS, and a high-accuracy IMU. The Raven will have two on-board cameras, a side-looking and a forward-looking optical camera. The Supervisor OCU will act as the central mission planner, allowing the operator to monitor mission events and override vehicle tasks.