Visible-Near-Infrared Scanners enable a noninvasive prediction of quality properties of fruit and vegetable based on previously created models. A combination of NIR scanners and
machine learning methods can lead to economic improvements and reduction of food waste by strategies like "first expired, first out" and dynamic pricing. In order to identify parameters
capable of showing dynamic postharvest development, three horticultural products with different postharvest behavior (e. g. strawberry, table grape and mango) were chosen for
morphological and statictical analysis. According to the results, a graduation of spectra in correspondence to the day of measurement was noticeable for strawberry regarding the a-
value as well as presumingly mass loss for both mango and table grape. Furthermore, a PLS model for the a-values r2cv = 0.80 was developed for strawberries.

Regulated deficit irrigation (RDI) comprises controlled reduction of irrigation during certain phases of plant development, accepting minor yield reduction but maximising net returns. To optimise RDI, plant response-based irrigation scheduling may be essential. For this, the knowledge of plant reactions to soil water regimes is indispensable. In field trials (randomized block design, 3 repetitions), ‘Nanthya’ carrots (F1-hybrid) were grown at 20, 40, 60 and 80% field capacity (determined by TDR sensors), yielding soil water contents of 5.1±0.2, 6.2±0.8, 10.1±0.5 and 11.2±0.2% (controls). To guarantee constant controlled soil water contents, plastic rain shelter (76% transparency) were applied. Plots were irrigated if necessary. Treatments started 23 d after sowing (DAS), first sampling was 91 DAS, while final harvest was 106 DAS. On each date, fresh and dry mass, water content, water potential, osmotic potential, turgor and stiffness of carrot tubers were analyzed. Root fresh mass (n=20 carrots per treatment) was higher at high water availability only at the 1st sampling but not at the final harvest. Here, water regime did not affect yield. At this time, root tuber water potential was lower than at the 1st sampling and it was also lower at low soil water availability (20%, 40%). Similar results were obtained for the mean osmotic potential. These variations partially resulted from osmotic adjustment (at 1st harvest) and from a lower root tuber water content at the final sampling. The generally higher modulus of elasticity at this date indicated water volume-independent cell wall-stiffening, i.e. elastic adjustment. The presented results point out that moderately reduced (by 20%) irrigation of carrots does not affect their yield, while pronounced water shortage may induce physiological adjustment.

Several lighting strategies in an indoor farm are compared regarding the annual electrical energy consumption. Based on measurements of the energy consumption during one experiment, conducted inside the indoor farm at the HSWT, an energy balancing model was calibrated and subsequently validated by a second experiment. The model was then used to calculate the potential reduction of the annual electrical energy consumption concerning several different lighting strategies. Furthermore, different settings of temperature set points and the impact of an optimized heat transfer coefficient as well as a more efficient performance for heating and cooling are considered. The measured and the modeled energy consumption and the values regained by the model showed a high regression coefficient (R2=0.936). The prediction of the energy consumption during the second experiment was also possible (R2=0.738) with a total difference to the measured consumption of 29.1 kWh. Regarding the given technical settings of the indoor farm, the annual electrical energy consumption can be reduced by up to 16% by an adjustment of the temperature strategy while reaching a similar yield. By assuming an optimized technical setting, the relevance of the lighting strategy increased significantly. Based upon this an annual reduction of the electrical energy consumption of up to 21% seemed conceivable.

Farmers and consultants face an unmanageable amount of diverse knowledge and information for crop management decisions. To determine optimal actions, decision makers require knowledge-based support. In this way, decisions can be improved and heuristics can be replaced over time. The study presents a digital knowledge base with an integrated decision support system (DSS), using the example of nutrient supply, specifically nitrogen (N), fertilization. Therefore, the requirements of farmers and crop consultants for DSS to inform fertilization decisions for winter wheat (Triticum aestivum L.) were elaborated using surveys, expert interviews, and a prototype test. Semantic knowledge was enriched by expert knowledge and combined in a web application, the Crop Portal. To map regional and personal decision making patterns and experiences, the tacit knowledge on the complex advisory problem of N fertilization is made digitally usable. For this purpose, 16 fuzzy variables were specified and formalized. Individual decision trees and their interactions with an integrative knowledge base were used to multiply the consulting reach of experts. Using three consultants and nine model farms from different soil–climate areas in Germany, the Crop Portal was tested under practical conditions and the perceived pragmatic and hedonic quality of the system was evaluated using a standardized questionnaire. The field test showed that the variation in fertilizer recommendations from the ‘digital advisor twin’ ranged from 5 kg N ha⁻¹ to 16 kg N ha⁻¹ when compared with the decisions of the experts in the field. The study presents the participatory development and evaluation of a rule-based DSS prototype in agricultural practice.