An omnidirectional multirotor has the advantageous maneuverability of decoupled translational and rotational motions, drastically superseding the traditional multirotors' motion capability. Such maneuverability requires an omnidirectional multirotor to frequently alter the thrust amplitude and even direction, which is prone to the rotors' settling time induced from the rotors' own dynamics. Furthermore, the omnidirectional multirotor's stability for tracking control in the presence of rotor dynamics has not yet been addressed. To resolve this issue, we propose a geometric tracking controller that takes the rotor dynamics into account. We show that the proposed controller yields the zero equilibrium of the error dynamics almost globally exponentially stable. The controller's tracking performance and stability are verified in simulations. Furthermore, the single-axis force experiment with the omnidirectional multirotor has been performed to confirm the proposed controller's performance in mitigating the rotors' settling time in the real world.
The performance of a robot controller depends on the choice of its parameters, which require careful tuning. In this paper, we present DiffTune, a novel, gradient-based automatic tuning framework. Our method unrolls the dynamical system and controller as a computational graph and updates the controller parameters through gradient-based optimization. Unlike the commonly used back-propagation scheme, the gradient in DiffTune is obtained through sensitivity propagation, a forward-mode auto differentiation technique that runs parallel to the system's evolution. We validate the proposed auto-tune approach on a Dubin's car and a quadrotor in challenging simulation environments. Simulation experiments show that the approach is robust to uncertainties in the system dynamics and environment and generalizes well to unseen trajectories in tuning.
Crop management, including nitrogen (N) fertilization and irrigation management, has a significant impact on the crop yield, economic profit, and the environment. Although management guidelines exist, it is challenging to find the optimal management practices given a specific planting environment and a crop. Previous work used reinforcement learning (RL) and crop simulators to solve the problem, but the trained policies either have limited performance or are not deployable in the real world. In this paper, we present an intelligent crop management system which optimizes the N fertilization and irrigation simultaneously via RL, imitation learning (IL), and crop simulations using the Decision Support System for Agrotechnology Transfer (DSSAT). We first use deep RL, in particular, deep Q-network, to train management policies that require all state information from the simulator as observations (denoted as full observation). We then invoke IL to train management policies that only need a limited amount of state information that can be readily obtained in the real world (denoted as partial observation) by mimicking the actions of the previously RL-trained policies under full observation. We conduct experiments on a case study using maize in Florida and compare trained policies with a maize management guideline in simulations. Our trained policies under both full and partial observations achieve better outcomes, resulting in a higher profit or a similar profit with a smaller environmental impact. Moreover, the partial-observation management policies are directly deployable in the real world as they use readily available information.
As Autonomous Vehicle (AV) development has progressed, concerns regarding the safety of passengers and agents in their environment have risen. Each real world traffic collision involving autonomously controlled vehicles has compounded this concern. Open source autonomous driving implementations show a software architecture with complex interdependent tasks, heavily reliant on machine learning and Deep Neural Networks (DNN), which are vulnerable to non deterministic faults and corner cases. These complex subsystems work together to fulfill the mission of the AV while also maintaining safety. Although significant improvements are being made towards increasing the empirical reliability and confidence in these systems, the inherent limitations of DNN verification create an, as yet, insurmountable challenge in providing deterministic safety guarantees in AV. We propose Synergistic Redundancy (SR), a safety architecture for complex cyber physical systems, like AV. SR provides verifiable safety guarantees against specific faults by decoupling the mission and safety tasks of the system. Simultaneous to independently fulfilling their primary roles, the partially functionally redundant mission and safety tasks are able to aid each other, synergistically improving the combined system. The synergistic safety layer uses only verifiable and logically analyzable software to fulfill its tasks. Close coordination with the mission layer allows easier and early detection of safety critical faults in the system. SR simplifies the mission layer's optimization goals and improves its design. SR provides safe deployment of high performance, although inherently unverifiable, machine learning software. In this work, we first present the design and features of the SR architecture and then evaluate the efficacy of the solution, focusing on the crucial problem of obstacle existence detection faults in AV.
Perception of obstacles remains a critical safety concern for autonomous vehicles. Real-world collisions have shown that the autonomy faults leading to fatal collisions originate from obstacle existence detection. Open source autonomous driving implementations show a perception pipeline with complex interdependent Deep Neural Networks. These networks are not fully verifiable, making them unsuitable for safety-critical tasks. In this work, we present a safety verification of an existing LiDAR based classical obstacle detection algorithm. We establish strict bounds on the capabilities of this obstacle detection algorithm. Given safety standards, such bounds allow for determining LiDAR sensor properties that would reliably satisfy the standards. Such analysis has as yet been unattainable for neural network based perception systems. We provide a rigorous analysis of the obstacle detection system with empirical results based on real-world sensor data.
Nitrogen (N) management is critical to sustain soil fertility and crop production while minimizing the negative environmental impact, but is challenging to optimize. This paper proposes an intelligent N management system using deep reinforcement learning (RL) and crop simulations with Decision Support System for Agrotechnology Transfer (DSSAT). We first formulate the N management problem as an RL problem. We then train management policies with deep Q-network and soft actor-critic algorithms, and the Gym-DSSAT interface that allows for daily interactions between the simulated crop environment and RL agents. According to the experiments on the maize crop in both Iowa and Florida in the US, our RL-trained policies outperform previous empirical methods by achieving higher or similar yield while using less fertilizers
The use of random sampling in decision-making and control has become popular with the ease of access to graphic processing units that can generate and calculate multiple random trajectories for real-time robotic applications. In contrast to sequential optimization, the sampling-based method can take advantage of parallel computing to maintain constant control loop frequencies. Inspired by its wide applicability in robotic applications, we calculate a sampling complexity result applicable to general nonlinear systems considered in the path integral method, which is a sampling-based method. The result determines the required number of samples to satisfy the given error bounds of the estimated control signal from the optimal value with the predefined risk probability. The sampling complexity result shows that the variance of the estimated control value is upper-bounded in terms of the expectation of the cost. Then we apply the result to a linear time-varying dynamical system with quadratic cost and an indicator function cost to avoid constraint sets.
This paper presents an approach for trajectory-centric learning control based on contraction metrics and disturbance estimation for nonlinear systems subject to matched uncertainties. The approach allows for the use of a broad class of model learning tools including deep neural networks to learn uncertain dynamics while still providing guarantees of transient tracking performance throughout the learning phase, including the special case of no learning. Within the proposed approach, a disturbance estimation law is proposed to estimate the pointwise value of the uncertainty, with pre-computable estimation error bounds (EEBs). The learned dynamics, the estimated disturbances, and the EEBs are then incorporated in a robust Riemannian energy condition to compute the control law that guarantees exponential convergence of actual trajectories to desired ones throughout the learning phase, even when the learned model is poor. On the other hand, with improved accuracy, the learned model can be incorporated in a high-level planner to plan better trajectories with improved performance, e.g., lower energy consumption and shorter travel time. The proposed framework is validated on a planar quadrotor navigation example.
A reinforcement learning (RL) policy trained in a nominal environment could fail in a new/perturbed environment due to the existence of dynamic variations. Existing robust methods try to obtain a fixed policy for all envisioned dynamic variation scenarios through robust or adversarial training. These methods could lead to conservative performance due to emphasis on the worst case, and often involve tedious modifications to the training environment. We propose an approach to robustifying a pre-trained non-robust RL policy with $\mathcal{L}_1$ adaptive control. Leveraging the capability of an $\mathcal{L}_1$ control law in the fast estimation of and active compensation for dynamic variations, our approach can significantly improve the robustness of an RL policy trained in a standard (i.e., non-robust) way, either in a simulator or in the real world. Numerical experiments are provided to validate the efficacy of the proposed approach.
Advances in remote sensing technology have led to the capture of massive amounts of data. Increased image resolution, more frequent revisit times, and additional spectral channels have created an explosion in the amount of data that is available to provide analyses and intelligence across domains, including agriculture. However, the processing of this data comes with a cost in terms of computation time and money, both of which must be considered when the goal of an algorithm is to provide real-time intelligence to improve efficiencies. Specifically, we seek to identify nutrient deficient areas from remotely sensed data to alert farmers to regions that require attention; detection of nutrient deficient areas is a key task in precision agriculture as farmers must quickly respond to struggling areas to protect their harvests. Past methods have focused on pixel-level classification (i.e. semantic segmentation) of the field to achieve these tasks, often using deep learning models with tens-of-millions of parameters. In contrast, we propose a much lighter graph-based method to perform node-based classification. We first use Simple Linear Iterative Cluster (SLIC) to produce superpixels across the field. Then, to perform segmentation across the non-Euclidean domain of superpixels, we leverage a Graph Convolutional Neural Network (GCN). This model has 4-orders-of-magnitude fewer parameters than a CNN model and trains in a matter of minutes.