Learning State-Variable Relationships in POMCP: A Framework for Mobile Robots.

We address the problem of learning relationships on state variables in Partially Observable Markov Decision Processes (POMDPs) to improve planning performance. Specifically, we focus on Partially Observable Monte Carlo Planning (POMCP) and represent the acquired knowledge with a Markov Random Field (MRF). We propose, in particular, a method for learning these relationships on a robot as POMCP is used to plan future actions. Then, we present an algorithm that deals with cases in which the MRF is used on episodes having unlikely states with respect to the equality relationships represented by the MRF. Our approach acquires information from the agent's action outcomes to adapt online the MRF if a mismatch is detected between the MRF and the true state. We test this technique on two domains, rocksample, a standard rover exploration task, and a problem of velocity regulation in industrial mobile robotic platforms, showing that the MRF adaptation algorithm improves the planning performance with respect to the standard approach, which does not adapt the MRF online. Finally, a ROS-based architecture is proposed, which allows running the MRF learning, the MRF adaptation, and MRF usage in POMCP on real robotic platforms. In this case, we successfully tested the architecture on a Gazebo simulator of rocksample. A video of the experiments is available in the Supplementary Material, and the code of the ROS-based architecture is available online.

The Control of the American Leafhopper Erasmoneura vulnerata (Fitch) in European Vineyards: Impact of Synthetic and Natural Insecticides.

The American leafhopper Erasmoneura vulnerata, detected in Europe in the early 2000s, has recently become a pest in North-Italian vineyards. Infestations were recorded in organic and conventional vineyards despite the application of insecticides targeting other pests. Erasmoneura vulnerata completes three generations per year, and the second generation is frequently associated with large populations. The selection of appropriate active ingredients and the timing of their application is crucial for effective pest control. Field trials were carried out in Northeastern Italy, using a randomized design, to evaluate the impact of insecticides applied against other grapevine leafhoppers on E. vulnerata populations. The beginning of the second generation was selected as the best time for insecticide application. For natural products, two applications were planned. Among the selected insecticides, the most effective were acetamiprid, flupyradifurone and lambda-cyhalothrin. Regarding natural products, the most effective was kaolin which could represent an alternative to pyrethrins in organic vineyards. The identification of pest threshold levels and the evaluation of side effects of the most effective insecticides on key natural enemies occurring in vineyards are required.

Managing the vine mealybug, Planococcus ficus, through pheromone-mediated mating disruption.

The vine mealybug (VMB), Planococcus ficus (Hemiptera: Pseudococcidae), is a key insect pest of vineyards. While pheromone-based mating disruption (MD) has been successfully tested against a wide range of insect pests, knowledge about its efficacy against key mealybug species, such as P. ficus, is scarce. In this study, a novel MD product, Isonet® PF, was evaluated by testing 300, 400, and 500 dispensers/ha at four study sites located in Northern (Veneto) and Southern (Sicily) Italy. Experiments were carried out over 2 years by monitoring the mealybug populations in wine grape and table grape vineyards managed with and without the application of MD. Pheromone dispensers were periodically collected during the grapevine-growing season, extracted, and analyzed by GC-MS, to determine their pheromone content and the release in mg/ha/day. The results showed that use of the MD dispenser Isonet® PF reduced the percentage of VMB-infested bunches and the number of VMB specimens per bunch compared with the untreated controls. This was recorded over 2 years at all experimental sites. Differences in the incidence of infested bunches among the three tested rates of Isonet® PF were not detected. Overall, the results presented here contribute to optimizing the sex pheromone dosage used in MD control programs against VMB allowing a reduction of broad-spectrum insecticides currently employed to manage this important pest.