Jörg Steffen Dittrich

An Unmanned Helicopter for Autonomous Flights in Urban Terrain

This work summarizes a multi-disciplinary research project, focusing on key enabling techniques towards true autonomous flight of small, low flying VTOL UAVs. Research activities cover the flying testbed, a simulation and testing environment, as well as integrated components for onboard navigation, perception, planning and control. Promising results and feasibility demonstrations in flight tests underline the successful domain specific enhancements of approaches based on aeronautical engineering, computer science and mobile robotics. The current approaches pave the way towards further research in improved flight control performance and more system autonomy when a-priori mission uncertainties increase.

Autonomous Vision-Based Helicopter Flights Through Obstacle Gates

The challenge for unmanned aerial vehicles to sense and avoid obstacles becomes even harder if narrow passages have to be crossed. An approach to solve a mission scenario that tackles the problem of such narrow passages is presented here. The task is to fly an unmanned helicopter autonomously through a course with gates that are only slightly larger than the vehicle itself. A camera is installed on the vehicle to detect the gates. Using vehicle localization data from a navigation solution, camera alignment and global gate positions are estimated simultaneously. The presented algorithm calculates the desired target waypoints to fly through the gates. Furthermore, the paper presents a mission execution plan that instructs the vehicle to search for a gate, to fly through it after successful detection, and to search for a proceeding one. All algorithms are designed to run onboard the vehicle so that no interaction with the ground control station is necessary, making the vehicle completely autonomous. To develop and optimize algorithms, and to prove the correctness and accuracy of vision-based gate detection under real operational conditions, gate positions are searched in images taken from manual helicopter flights. Afterwards, the integration of visual sensing and mission control is proven. The paper presents results from full autonomous flight where the helicopter searches and flies through a gate without operator actions.

Mapping and path planning in complex environments: An obstacle avoidance approach for an unmanned helicopter

This paper presents an obstacle avoidance method that is performed with an unmanned helicopter. The approach begins with a mapping step where information from sensor data about previously unknown dangers is extracted into an occupancy grid and eventually converted into a polygonal 3D world model. This continuously updating map is used by a path planner that generates and updates a 3D trajectory guiding the vehicle through safe passages around the detected objects. The algorithms are generic but optimized for unmanned aircraft and a stereo camera as the environmental sensor. Computation is fully executed on board so that a ground control station is only needed for supervision. With successful obstacle detection and avoidance flight tests, the paper shows the qualification of the presented method under real operational conditions.

Adaptive Trajectory Controller for Generic Fixed-Wing Unmanned Aircraft

This work deals with the construction of a nonlinear adaptive trajectory controller, which is easily applicable to a multitude of fixed wing unmanned aircraft. Given a common signal interface, the adaptive trajectory controller is divided into a generic part, which is common for each vehicle, and into a part, which is unique. The generic part of the control architecture bases on a common inversion model which is used for feedback linearization. However, the dynamics of the aircraft and the inversion model differ, thus introducing model uncertainties to the feedback linearized system. The effect of modeling uncertainties is reduced by the application of a concurrent learning model reference adaptive controller, which uses neural networks in order to approximate the uncertainty. Leveraging instantaneous as well as stored data concurrently for adaptation ensures convergence of the adaptive parameters to a set of optimal weights, which minimize the approximation error. Performance and robustness against certain model uncertainties is shown through numerical simulation for two significantly different unmanned aircraft.

Aerial Tracking and GNSS-Reference Localization with a Robotic Total Station

Precise navigation for small unmanned aerial systems is of high interest for disaster scenarios and many other tasks. Advanced navigation technologies for future low-altitude applications in completely or partly unknown environments are carried out by DLRs Unmanned Aircraft Department in cooperation with the Institute of Flight Guidance from the Technical University Braunschweig. This paper describes the usage of a robotic total station as a reference localization system in order to benchmark current and developing satellite, inertial, and optical navigation techniques. Such devices perform laser-based distance measurements to the aircraft and track the moving vehicle. Since total stations are not mainly designed for aerial applications, the paper includes the description of some required extensions like faster automatic targeting and time synchronization. The paper concludes with the analysis of helicopter flight tests, where satellite-based localization results from different receivers are compared with the total station reference measurements.

A Flight State Estimator that Combines Stereo-Vision, INS, and Satellite Pseudo-Ranges

This paper presents a flight state estimator which couples stereo vision, inertial (INS), and global navigation satellite system (GNSS) data. The navigation filter comes with different operation modes that allow loosely coupled GNSS/INS positioning and, for difficult conditions, improvements using visual odometry and a tighter coupling with GNSS pseudo-range (PSR) data. While camera systems are typically used as an additional relative movement sensor to enable positioning without GNSS for a certain amount of time, the PSR data filtering allows to use satellite navigation also when less than four satellites are available. This makes the filter even more robust against temporary dropouts of the full GNSS solution. The application is the navigation of unmanned aircraft in disaster scenarios which includes flights close to ground in urban or mountainous areas. The filter performance is evaluated with sensor data from unmanned helicopter flight tests where different conditions of the GNSS signal reception are simulated. It is shown that the use of PSR data improves the positioning significantly compared to the dropout when the signals of less than four satellites are available.

Adapting Scenario Definition Language for Formalizing UAS Concept of Operations

Recent regulatory efforts from EASA introduced the new concept of a specific category that allows a stepwise adaptation of certification requirements. Based on a specific operation risk assessment, this category enables new aircraft system architectures and mission designs. The concept allows for operational restrictions, defined in the concept of operations, to maintain overall operation safety. Therefore, the description of the concept of operations is one of the key elements for the approval of a UAS operation. This description will be used by pilots, operators, certification authorities and possibly additional stakeholders, e.g., for publishing operation into databases. Standardization of this document seems mandatory. In addition to this, we discuss the formalization of this document for easy distribution amongst stakeholders and establishment of tool support. To support this, we use recent developments regarding standardization of simulation modeling. The American Institute of Aeronautics and Astronautics Modeling and Simulation Technical Committee recently launched a working group towards the development of a standard scenario definition language for aviation. In this paper, we extend this language specification in order to define specific operation safety boundaries of a system. Thereby, we aim at creating a dialect of the simulation scenario definition language that can be used to describe safe scenarios for a particular system of interest. We further discuss the challenges and benefits of such a formalization of the concept of operations by adapting the simulation scenario definition language.