The development of agricultural robots is an increasingly popular research field aiming at addressing the widespread labor shortages in the farming industry and the ever-increasing food production demands. In many cases, multiple cooperating robots can be deployed in order to reduce task duration, perform an operation not possible with a single robot, or perform an operation more effectively. Building on previous results, this application paper deals with a cooperation strategy that allows two heterogeneous robots to cooperatively carry out grape harvesting, and its implementation is demonstrated. More specifically, the cooperative grape harvesting task involves two heterogeneous robots, where one robot (i.e., the expert) is assigned the grape harvesting task, whereas the second robot (i.e., the helper) is tasked with supporting the harvesting task by carrying the harvested grapes. The proposed cooperative harvesting methodology ensures safe and effective interactions between the robots. Field experiments have been conducted in order firstly to validate the effectiveness of the coordinated navigation algorithm and secondly to demonstrate the proposed cooperative harvesting method. The paper reports on the conclusions drawn from the field experiments, and recommendations for future enhancements are made. The potential of sophisticated as well as explainable decision-making based on logic for enhancing the cooperation of autonomous robots in agricultural applications is discussed in the context of mathematical lattice theory.

Sustainable agricultural production, under the current world population explosion, calls for agricultural robot operations that are personalized, i.e., locally adjusted, rather than en masse. This work proposes implementing such operations based on logic in order to ensure that a reasonable operation is applied locally. In particular, the interest here is in grape harvesting, where a binary decision has to be taken regarding the maturity of a grape in order to harvest it or not. A Boolean lattice ontology of inequalities is considered regarding three grape maturity indices. Then, the established fuzzy lattice reasoning (FLR) is applied by the FLRule method. Comparative experimental results on real-world data demonstrate a good maturity prediction. Other advantages of the proposed method include being parametrically tunable, as well as exhibiting explainable decision-making with either crisp or ambiguous input measurements. New mathematical results are also presented.

The all-popular Artificial Intelligence (AI) is pursued by deep learning (neural computing) models implemented in Data Centers (DCs) that require considerable resources. The aforementioned models are typically developed in the Euclidean (Hilbert) space R^N, where R is the set of real numbers for an integer N. This work introduces a Hibert space, namely, space of Generalized Intervals’ Numbers (GINs) G as a hierarchy of Hilbert spaces stemming from R. Data processing is carried out in the convex cone F₁ of G, namely, space of Intervals’ Numbers (INs), which includes partially ordered, Lebesque space L₂ distribution functions that may represent information granules, for example, probability/possibility distributions. It is detailed how enhanced deep learning models can be developed in F1N. As F₁ is a strict superset of R, all the conventional deep learning algorithms in R^N are included. Additional advantages of deep learning in F1N include (a) accommodation of spoken language semantics, represented by partial order, (b) potential engagement of axiomatic logic all along during data processing, and (c) tuning the number of tunable parameters toward reducing the demands for DC resources, including energy consumption, by engaging fewer models of greater flexibility without decreasing performance.