James Goppert

Cyber Attack Vulnerabilities Analysis for Unmanned Aerial Vehicles.

As the technological capabilities of automated systems have increased, the use of unmanned aerial vehicles (UAVs) for traditionally exhausting and dangerous manned missions has become more feasible. The United States Army, Air Force, and Navy have released plans for the increased use of UAVs, but have only recently shown interest in the cyber security aspect of UAVs. As a result, current autopilot systems were not built with cyber security considerations taken into account, and are thus vulnerable to cyber attack. Since UAVs rely heavily on their on-board autopilots to function, it is important to develop an autopilot system that is robust to possible cyber attacks. In order to develop a cyber-secure autopilot architecture, we have run a study on potential cyber threats and vulnerabilities of the current autopilot systems. This study involved a literature review on general cyber attack methods and on networked systems, which we used to identify the possible threats and vulnerabilities of the current autopilot system. We then studied the identi ed threats and vulnerabilities in order to analyze the post-attack behavior of the autopilot system through simulation. The uses of UAVs are increasing in many applications other than the traditional military use. We describe several example scenarios involving cyber attacks that demonstrate the vulnerabilities of current autopilot systems.

Model Checking of a Flapping-Wing Mirco-Air-Vehicle Trajectory Tracking Controller Subject to Disturbances

This paper proposes a model checking method for a trajectory tracking controller for a flapping wing micro-air-vehicle (MAV) under disturbance. Due to the coupling of the continuous vehicle dynamics and the discrete guidance laws, the system is a hybrid system. Existing hybrid model checkers approximate the model by partitioning the continuous state space into invariant regions (flow pipes) through the use of reachable set computations. There are currently no efficient methods for accounting for unknown disturbances to the system. Neglecting disturbances for the trajectory tracking problem underestimates the reachable set and can fail to detect when the system would reach an unsafe condition. For linear systems, we propose the use of the H-infinity norm to augment the flow pipes and account for disturbances. We show that dynamic inversion can be coupled with our method to address the nonlinearities in the flapping-wing control system.

Numerical Analysis of Cyberattacks on Unmanned Aerial Systems.

Cyber security has emerged as one of the most important issues in the operation of Unmanned Aerial Systems (UASs) due to their heavy reliance on the on-board autopilot systems. As the first step to implement an autopilot system that is robust to possible cyber attacks, we have conducted a study to identify potential cyber threats and vulnerabilities inherent in the given UASs. In an attempt to study these vulnerabilities in more detail, this paper presents an analytical algorithm to test the behavior of UASs under various cyber attacks and quantify their severity accordingly. Compared to a numerical approach, the analytical algorithm enables the prediction of the most effective cyber attack combinations without the need to compute the severity of all the attack combinations, thereby greatly reducing the computational cost. The performance of the proposed algorithm is demonstrated with a linearized longitudinal motion of a rotorcraft example.

Realization of an Autonomous, Air-to-Air Counter Unmanned Aerial System (CUAS)

The proliferation of small Unmanned Aerial Systems (UASs) has led to a security gap in the defense of strategic installations and at public events. One of the most proven and low-regret methods employed by Counter Unmanned Aerial Systems (CUASs) is entanglement of the hostile UAS in a net carried by a hunter UAS. Typically these hunter UASs are controlled by a human pilot. We employ a ground based RADAR system for tracking the target UAS and command the hunter UAS to follow the target UAS, using the Robot Operating System (ROS), the MAVLink protocol, and the PX4 autopilot. This system is fully autonomous, which reduces cost and response time when compared to human-in-the-loop systems. In addition, a novel cylindrical net design is presented. We demonstrate the systems effectiveness through field testing.

Model Checking of a Training System Using NuSMV for Humanoid Robot Soccer

Model checking is a technique to perform a formal verification process that allows a system to have robustness and correctness. In a given system model as a Finite State Machine (FSM), model checker explores all possible states in brute-force manner. In this paper, we apply this technique to a training system, which teaches a humanoid soccer robot how to intercept a ball that is passed from other players, to verify that the system is failure-safe in a given requirements. Several Computation Tree Logic (CTL) properties to define a critical or potential situation are specified based on the functionality of the system. We show the results of the given properties using NuSMV, a symbolic model checker introduced by Carnegie Mellon University.

Optimizing Energy Efficiency of a Flapping Robotic Bird Through Application of Evolutionary Algorithms

Recent advances in servo motors have enabled construction of a new low cost class of flapping wing robots. These robotic, bird-sized flapping wing vehicles are capable of independently controlling wing position. In contrast, most previously-developed flapping wing mechanisms rely on mechanical linkages to oscillate the wings and can only control the flapping frequency. This new class of servo-based robotic birds allows for minimization of the number of necessary actuators and can provide for more degrees of freedom, eliminating the need for additional control surfaces. As with many other flapping wing vehicles, power consumption is of great concern, since it is the limiting operational factor of flight duration. In this paper, we investigate the use of evolutionary algorithms to optimize the flapping and gliding patterns of a robotic bird. For gliding, we maximize the lift to drag ratio while minimizing the aerodynamic moments and side force. For flapping, we maximize the lift while minimizing the drag and power consumption by modulating the frequency of the up and down wing strokes. In order to allow on-board learning, we modify an existing design to have similar aero-surface area and significantly reduced weight, accommodating an increased computational payload.

Constructing BADA Models of Existing UASs from Public Data and Photos using Open Source Tools

The commercial and military use of unmanned aerial systems (UASs) continues to grow and thus safe integration of unmanned and manned aircraft in the same airspace will require detailed knowledge of UASs flight performance. The current standard for information used in air traffic control analysis is BADA, an aircraft performance model developed by Eurocontrol. However, due to corporate and government confidentiality, BADA models are not publicly available for current UASs. We present a methodology for employing publicly available data and open source software tools to generate this high fidelity model and extract BADA data from the model through static analysis of the aircraft’s flight performance. To analyze the accuracy of the proposed method, we conduct our analysis on a small, commercially available UAS and compare the estimated flight performance using our method with results obtained through wind tunnel testing. The accuracy of the models produced by the proposed method depends strongly on the amount of data available and the similarity of the system to existing systems with known characteristics. Our method is shown to be an effective strategy for estimating the flight performance and deriving models of UASs with limited public data.

Invariant Kalman filter application to optical flow based visual odometry for UAVs

Optical flow based visual odometry for UAVs has become akin to wheel encoders for ground based robots. While sensors such as laser rangefinders and Global Positioning System (GPS) receivers can provide measurements of a UAVs position, these sensors typically have a low bandwidth and can become degraded (e.g. GPS in urban canyons). Optical flow sensors provide a robust high bandwidth pseudo-velocity measurement by tracking the movement of a feature through a camera image and measuring the distance to that feature, typically using a sonar or a lidar sensor. Optical flow based visual odometry thus compliments low bandwidth UAV position measurements. We have previously used a simple linear measurement equation to approximate the optical flow as a pseudo-velocity measurement and were able to achieve fully autonomous mission flights without GPS both indoors and outdoors. This estimator, known as Local Position Estimator (LPE), is now part of the open source PX4 autopilot. In this work, we seek to improve the UAVs performance in terms of maximum speed and robustness by deriving an estimator using the full nonlinear measurement equations and by basing the estimator on the Invariant Extended Kalman Filter (IEKF). Through intelligent choice of the frame in which the estimator dynamics and measurement equations are linearized, the IEKF is able to reduce the fluctuations in the Kalman filter along typical vehicle trajectories and produce a more optimal estimate. We compare our previous algorithm, LPE, with our new algorithm, IEKF, using the PX4 gazebo based software in the loop simulator.

A Provisional Approach to Maintaining Verification and Validation Capability in Self-Adapting Robots

Cyber Physical Systems (CPS) are composed of multiple physical and computing components that are deeply intertwined, operate on differing spatial and temporal scales, and interact with one another in fluid, context dependent, manners. Cyber Physical Systems often include smart components that use local adaptation to improve whole system performance or to provide damage response. Evolvable and Adaptive Hardware (EAH) components, at least conceptually, are often represented as an enabling technology for such smart components. This paper will outline one approach to applying CPS thinking to better address a growing need to address Verification and Validation (V&V) questions related to the use of EAH smart components. It will argue that, perhaps fortuitously, the very adaptations EAH smart components employ for performance improvement may also be employed to maintain V&V capability.

Using Online Model Checking Technique for Survivability, Evaluating Different Scenarios on Runtime.

Abstract Survivability or detecting and predicting failures for humans or animals is a matter of instinct. Unlike for robots, machines, or anything that bases its rationality in software, that assumption is not applicable, yet. In this paper we apply model checking techniques to avoid future complete malfunctioning. Scenarios where non existing states are evaluated in order to find solution to possible future problems. A model checker is used for online evaluation of all the states, including possible non-existing transitions and states.