The use of deep neural networks (DNNs) in plant phenotyping has recently received considerable attention. By using DNNs, valuable insights into plant traits can be readily achieved. While these networks have made considerable advances in plant phenotyping, the results are processed too slowly to allow for real-time decision-making. Therefore, being able to perform plant phenotyping computations in real-time has become a critical part of precision agriculture and agricultural informatics. In this work, we utilize state-of-the-art object detection networks to accurately detect, count, and localize plant leaves in real-time. Our work includes the creation of an annotated dataset of Arabidopsis plants captured using Cannon Rebel XS camera. These images and annotations have been complied and made publicly available. This dataset is then fed into a Tiny-YOLOv3 network for training. The Tiny-YOLOv3 network is then able to converge and accurately perform real-time localization and counting of the leaves. We also create a simple robotics platform based on an Android phone and iRobot create2 to demonstrate the real-time capabilities of the network in the greenhouse. Additionally, a performance comparison is conducted between Tiny-YOLOv3 and Faster R-CNN. Unlike Tiny-YOLOv3, which is a single network that does localization and identification in a single pass, the Faster R-CNN network requires two steps to do localization and identification. While with Tiny-YOLOv3, inference time, F1 Score, and false positive rate (FPR) are improved compared to Faster R-CNN, other measures such as difference in count (DiC) and AP are worsened. Specifically, for our implementation of Tiny-YOLOv3, the inference time is under 0.01 s, the F1 Score is over 0.94, and the FPR is around 24%. Last, transfer learning using Tiny-YOLOv3 to detect larger leaves on a model trained only on smaller leaves is implemented. The main contributions of the paper are in creating dataset (shared with the research community), as well as the trained Tiny-YOLOv3 network for leaf localization and counting.

Islanded nanogrids (NGs) are autonomous systems consisting of small-scale generation units including renewable energy sources and traditional fuel generators and energy storage systems (ESS) that typically serve few buildings or loads. This work aims at developing and validating a new optimal energy management (EM) algorithm for an islanded NG. To minimize the generator's operating cost and maximize battery availability at each operating cycle, dynamic programming (DP) framework is employed to solve the underlying optimization problem. The goal of the proposed approach is to ensure the use of maximum available solar power and to achieve optimal battery state of charge (SOC). To meet that goal, the management of the ESS is formulated as a stochastic optimal control problem, where nonlinearities in the battery discharging process are considered. A Markov model is constructed for predicting the probability distribution of the solar production used in the stochastic DP formulation. Simulation results are given to illustrate the efficacy of the proposed DP-based approach compared to a rule-based algorithm. Finally, a hardware-in-the-loop system is used to evaluate the real-time operation of the proposed EM algorithm.

This paper presents a stochastic and predictive control design approach for connected and automated vehicles (CAVs) in a mixed-autonomy traffic environment, where CAVs are able to react properly to uncertain maneuvers of human-driven vehicles (HVs). The proposed fully-automated cooperative adaptive cruise control (CACC) design leverages a discrete hybrid stochastic model predictive controller that automatically determines the vehicle's operating mode based on onboard sensors data and information received through vehicle-to-vehicle (V2V) communication. Operating modes include free following, warning, danger, emergency braking, and lane change. Although the controller mainly focuses on maintaining the desired velocity and distance among CAVs, it also allows HVs to perform lane-change maneuvers and merge into the platoon's lane when needed. In response to an HV's position in the lane and its probabilistic behavior, the controller may switch the CAV's operating mode to react accordingly. Considering free-following and emergency-braking modes leads to efficient and safe autonomous driving. Switching between warning, danger, and lane-change modes along with adjusting the steering angle to perform a lane-change maneuver, when needed, robustifies the platoon's performance against unexpected human-driven vehicle maneuvers. Simulation studies are conducted to validate the efficacy of the proposed control design approach. The performance of the proposed control design approach is also compared to a switching control using simulation studies.

Purpose: This paper aims to focus on the design and testing of a robotic device for power line inspection and cleaning. The focus for this design is on simplicity and compactness with a goal to create a device for linemen and other power line workers to keep in their toolbox. Design/methodology/approach: The prototype uses V-grooved wheels to grip the line and can pass obstacles such as splices. It is equipped with a video camera to aid in line inspection and a scrub brush to clean debris from the line. The operator controls the device remotely from a laptop through a wireless connection. The novel way in which this device moves down the power line allows compactness while still being able to overcome in-line obstacles up to a certain size. Findings: The device has been tested on a test bed in the lab. The device is able to move down a line and expand to overcome in-line obstacles as it travels. Testing proved the mechanical feasibility and revealed new requirements for a future prototype. Practical implications: The device can be used for power line asset management by power companies; line inspection can lead to preventative repairs, leading to less downtime. Social implications: It stands to reduce costs related to maintenance and mitigates down time and emergency repairs. Originality/value: Innovative features include its size, mobility and control methods. Overall, the impact of this work extends to the utility maintenance sector and beyond.

The potential impacts of data integrity attacks on multi‐settlement electricity markets have been recently investigated and have sent a strong message to power grids independent system operators (ISOs) that adversaries could launch profitable cyber attacks by casting an incorrect image of transmission lines congestion pattern. However, these cautionary messages may be underestimated due to the adversaries unrealistic requirements (e.g. having access to real‐time measurements) to launch a successful stealthy and profitable attack. This study examines the potential of the aforementioned risk by demonstrating how a malicious power market participant could disturb the electricity market operation, using a pre‐designed false data injection attack along with bogus electricity trades in both day‐ahead and real‐time markets. The proposed attack design is robust against market uncertainties and the adversary can guarantee the success of the attack in advance. Hence, the existence of such cyber attacks against electricity markets can make the adversaries more aggressive. The numerical results on the IEEE 14‐bus test system confirm the vulnerability of multi‐settlement electricity markets to such financial cyber attacks. The results obtained from investigating such an attack design can be employed by ISOs in order to provide appropriate countermeasures.