The paper presents a working pipeline which integrates hardware and software in an automated robotic rose cutter. To the best of our knowledge, this is the first robot able to prune rose bushes in a natural environment. Unlike similar approaches like tree stem cutting, the proposed method does not require to scan the full plant, have multiple cameras around the bush, or assume that a stem does not move. It relies on a single stereo camera mounted on the end-effector of the robot and real-time visual servoing to navigate to the desired cutting location on the stem. The evaluation of the whole pipeline shows a good performance in a garden with unconstrained conditions, where finding and approaching a specific location on a stem is challenging due to occlusions caused by other stems and dynamic changes caused by the wind.

This paper presents the development, testing and validation of SWEEPER, a robot for harvesting sweet pepper fruit in greenhouses. The robotic system includes a six degrees of freedom industrial arm equipped with a specially designed end effector, RGB‐D camera, high‐end computer with graphics processing unit, programmable logic controllers, other electronic equipment, and a small container to store harvested fruit. All is mounted on a cart that autonomously drives on pipe rails and concrete floor in the end‐user environment. The overall operation of the harvesting robot is described along with details of the algorithms for fruit detection and localization, grasp pose estimation, and motion control. The main contributions of this paper are the integrated system design and its validation and extensive field testing in a commercial greenhouse for different varieties and growing conditions. A total of 262 fruits were involved in a 4‐week long testing period. The average cycle time to harvest a fruit was 24 s. Logistics took approximately 50% of this time (7.8 s for discharge of fruit and 4.7 s for platform movements). Laboratory experiments have proven that the cycle time can be reduced to 15 s by running the robot manipulator at a higher speed. The harvest success rates were 61% for the best fit crop conditions and 18% in current crop conditions. This reveals the importance of finding the best fit crop conditions and crop varieties for successful robotic harvesting. The SWEEPER robot is the first sweet pepper harvesting robot to demonstrate this kind of performance in a commercial greenhouse.

This article describes the tasks and first results of the work package "Manipulator and Control" of the EU project Trimbot2020. This project develops a mobile robot for outdoor hedge, rose and bush trimming. The Kinova Jaco(2) robotic arm was selected as manipulator. Two different types of robotic end-effectors have been developed. The tool for trimming topiaries uses two custom designed circular contra-rotating blades. The tool for single stem cutting is based on a commercial electrical pruner. The arm and the tools can all be controlled by using the Robot Operating System (ROS). The motion planning algorithm of the arm for the bush trimming action is divided into the planning setup module, the coverage planning module and the trajectory planning module. The path planning is modelled as a traveling salesman problem. In the first phase of the project the trimming control is performed open loop. A positioning genetic algorithm was developed that minimizes the needed number of vehicle poses for one target object. In the next phase of the project a vision feedback mechanism will be implemented.

This paper evaluates a robot developed for autonomous harvesting of sweet peppers in a commercial greenhouse. Objectives were to assess robot performance under unmodified and simplified crop conditions, using two types of end effectors (Fin Ray; Lip type), and to evaluate the performance contribution of stem‐dependent determination of the grasp pose. We describe and discuss the performance of hardware and software components developed for fruit harvesting in a complex environment that includes lighting variation, occlusions, and densely spaced obstacles. After simplifying the crop, harvest success significantly improved from 6% to 26% (Fin Ray) and from 2% to 33% (Lip type). We observed a decrease in stem damage and an increase in grasp success after enabling stem‐dependent determination of the grasp pose. Generally, the robot had difficulty in successfully picking sweet peppers and we discuss possible causes. The robot's novel capability of perceiving the stem of a plant may serve as useful functionality for future robots.

This paper focusses on field experiments with two different types of end-effectors for robotic harvesting of sweet-pepper fruits. One of the major issues is to reach, grasp and detach the fruit efficiently, without damaging it, while avoiding obstacles in the environment. End-effectors for harvesting fruit must be able to adapt to different fruit sizes and geometries. Two types of end-effectors were designed and realized. The first one had four fingers which utilized the "Fin Ray" effect to grip the fruit. A scissor-like cut mechanism on top of the fingers was used to cut through the fruit peduncle. The second, a lip-type end-effector first stabilized the fruit using a suction cup after which two rings enclosed and cut the peduncle with a circular blade integrated in the upper lip. Both end-effectors had integrated miniature cameras with a LED illumination system: one Time of Flight camera and the other a colour camera. To study the performance of the end-effectors a number of harvesting experiments were performed in commercial sweet-pepper greenhouses. Special attention was paid to the following aspects: positioning at the target fruit, separation of the fruit from the plant, fruit damage, leaf damage and plant stem damage. Both end-effector designs had their strengths and weaknesses. The Fin ray type end-effector harvested a maximum of 80% of the fruits on the plant, the lip-type end-effector a maximum of 76% of the fruits. In none of the experiments more than 64% of the fruit could be harvested without fruit damage.

Agricultural environments impose high demands on robotic grippers since the objects to be grasped (e.g., fruit) suffer from inherent uncertainties in size, shape, weight, and texture, are typically highly sensitive to excessive force, and tend to be partly or fully occluded. This paper presents a methodology for evaluating the influence of perception capabilities on grasping and on gripper design using graspability maps. Graspability maps are spatial representations of grasp quality grades from wrist poses (position and orientation) about an object and are generated using simulation. A new module was developed to enable the insertion of object pose errors for testing the effects of perception inaccuracies on grasping. The methodology was implemented for comparing two grippers (Fin-Ray and Lip-type) for harvesting two sweet-pepper cultivars. A 3D model of each gripper was constructed and suitable grasp quality measures were developed and validated in a physical environment. Task and gripper specific grasp quality measures were developed for each implementation. Sensitivity analyses included varying pepper dimensions and perception inaccuracies. These were followed by analyses of the influence of gripper design parameters on grasp capabilities. Results indicate that the Lip-type gripper is less sensitive to inaccuracies in object orientation, while both grippers are similarly sensitive to inaccuracies in object position. Specific perception system demands and design recommendations are given for each gripper, and cultivar. The results illustrate the importance of integrating perception analysis in the gripper design phase and the utility of the graspability simulation tool for design analysis. (C) 2015 IAgrE. Published by Elsevier Ltd. All rights reserved.

This paper describes the results of the development of a robot for harvesting sweet-peppers in greenhouses. A description is given of the working environment of the robot and its design objectives. The base of the robot consists out of two carrier modules. On the first, the manipulator, the control electronics and the computers are located. To assure maximum flexibility the realized manipulator prototype has nine degrees-of-freedom. On the second, the sensors and illumination are placed. The coupled modules can move in between the crop rows on the greenhouse rail system. The heights of the modules can be adjusted to match the height of the crop. On the sensor carrier module two 5 megapixel colour cameras and a Time of Flight camera are installed. The colour images and three dimensional (3D) data were calibrated and registered. Around the sensors, a lighting rig is placed to illuminate the scene. The sensor system is mounted on a linear motorized slide and can be horizontally moved in and out of the workspace of the manipulator. Machine vision software localises ripe fruits and obstacles in 3D. For fruit detection different approaches have been developed. One option is to initiate fruit detection by simple red colour blob detection. Another option is to perform fruit localization in two sequential steps. First Regions of Interest in the RGB image are selected which is suspected to contain target fruits. Next the fruit localization is performed in the corresponding 3D data, based on 3D point cloud template matching. Obstacle detection algorithms are used to localise plant stems and non-target fruits using the small baseline stereo images. In order to harvest the fruits, a motion planning module assures a collision free path for the manipulator to position the end-effector at the harvesting position. Two different types of end-effectors were designed and tested. The “Fin-Ray gripper” features a combined grip and cut mechanism. This end-effector first grips the fruit and after that the peduncle of the fruit is cut. The “Lip-type end-effector”, first stabilizes the fruit using a suction cup after which two rings enclose the fruit and cut the peduncle of the fruit. Both end effectors have a miniature RGB and a ToF camera for refining the fruit position and to determine the fruit pose. The main software platform of the robot was implemented for the Linux operating system and uses the open source middleware Robot Operating System (ROS). The coordinating control structure is based on a finite state machine and includes diagnostic tools and performance measures. The system was tested under simplified laboratory conditions in 2013. During these early tests 189 out of 194 fruit could be detected (97%), 167 fruits could be reached (86% of all fruits) and 154 picked (79% of all fruits). In spring 2014 final system integration took place. In a commercial greenhouse it was proven that the system is able to harvest pepper fruits fully autonomously.

One of the main targets of a greenhouse control system is to generate a uniform climate environment, which has obvious economic advantages, due to homogenous crop, lesser diseases and the potential energy saving. By using a dense wireless sensor network, the spatial distribution of climate parameters and thus the uniformity of greenhouse environment can be determined. This information can be used by the growers to detect cold and wet spots in greenhouse environment and to adjust their climate control system. In this study, the horizontal climate heterogeneity of the greenhouse environment is analysed. Six experiments were performed in commercial greenhouses observing the micro-climate with 100 wireless sensors under different conditions and for different crops. From the data the spatial and temporal distribution of air temperature T and relative humidity RH was obtained. Instantaneous spatial differences varied up to +/- 5 degrees C for T and +/- 20% for RH, depending on location, time of day and season. Significant differences of +/- 2 degrees C for T and +/- 12.5% for RH, while averaging data over longer periods (5 days), have been observed. Further research is needed to investigate the way the information regarding the spatial and temporal distribution of climatic parameters can be integrated in the control systems and finally to quantify the energy saving potential.

In buitenteelten, die uit de grond in substraat of een recirculerende waterlaag worden geteeld, valt meer regenwater dan de gewassen op jaarbasis nodig hebben. Het neerslagoverschot komt afhankelijk van het gewas in het teeltsysteem terecht en wordt dan vervuild met nutrienten en mogelijk gewasbeschermingsmiddelen geloosd. De grootte van de mogelijk te lozen stroom is onbekend en teeltsysteemafhankelijk. In deze studie is een overzicht gemaakt voor de verschillende teeltsystemen die bij “Teelt de Grond Uit” worden toegepast en wat daar de gevolgen van zijn als het neerslagoverschot moet worden geloosd en hoe groot die stroom is op jaarbasis.

In previous research a new type of greenhouse with an integrated concentrated photovoltaic system (CPV) was developed based on a circular covering geometry and an integrated filter for reflecting the near infrared radiation (NIR) of the greenhouse and exploiting this radiation in a solar energy system. The performance of the system was promising. In this study further optimalisation of the CPV system is made to avoid the large construction for solar tracing and the high investment. Hereto all parts for the solar concentrating system will be integrated into the greenhouse. The NIR-reflector material is carried out as a NIR-reflective lamellae system and the CPV-module is mounted into the ridge. In this paper the results of the optimization process of the CPV system based on NIR reflecting lamellae is presented. The optimization process is based on a maximal total annual electricity production and is performed with a ray tracing model and actual radiation data. Results show that the optimization of the lamellae greenhouse can be seen from a theoretical and a practical point of view. Theoretically, the number of lamellae for the investigated concept must be high (>100) and focus with a generic focal length of 3.5 m and glazing bars must be avoided. Then the maximal annual electricity output can be over 26 kWh/m(2). In practice, mechanical restrictions, plant conditions and costs will determine the implementation. The proposed CPV-system has positive side-effects like reducing the heat load (and the need for cooling) during summer and blocking of the direct radiation which can be harmful for some crops. With this, the feasibility of the system depends greatly on local conditions which require a tailor-made economic analysis.

In this paper the design and development of a new type of greenhouse with an integrated filter for reflecting near infrared radiation (NIR) and a solar energy delivery system is described. Especially the optical parts as the spectral selective film, the properties of the circular reflector and the efficiencies of photo voltaic cells are studied. As a first measure, the spectral selective cover material, which prevents the entrance of NIR radiation, is investigated. It has to block up to 35% of the solar energy outside the greenhouse, which will reduce the needed cooling capacity. The second measure is the integration with a solar energy system. When the NIR reflecting coating is designed as a circular shaped reflector integrated in the greenhouse, the reflected solar energy of a PhotoVoltaic (PV) cell in the focus point delivers electric energy. With a ray tracing computer program the optimal geometry of the reflector was designed with respect to the collecting efficiency. The PV cells mounted in the focal point require cooling due to the high heat load of the concentrated radiation (geometric concentration factor of 30). The properties of different PV materials were investigated to find the optimal cell for this application. Cooled greenhouses are an important issue to cope with the combination of high global radiation and high outdoor temperatures. All parts are integrated in a 100m2 prototype greenhouse which will be applied for the proof of principle.

A greenhouse with Fresnel lenses in the south facing roof and a receiver for concentrated Photovoltaics with water cooling (CPVT system) will result in electrical and thermal energy output from the solar energy excess entering a greenhouse. The PV system converts about half of the direct radiation into heat and electricity. During periods with direct radiation this will significantly reduce the heat load on the greenhouse. For an optimal performance the roof elements must be asymmetric with a steep inclination at the north side (the exact angle of course depends on the latitude of the building site). The Fresnel lens structure is best oriented in upwards direction. In the current design, two lenses are placed in the inner space of a double glass. This prevents pollution and condensation on the lenses. By the upward facing of the lens structure, the focus quality is preserved over a much broader range of angles of incidence compared to a lens with downward facing structures. Each PMMA lens with a size of 1.20x1.52 m is composed of 12 'tiles' for easy production. The focal distance of the lens is 1,875 m and the geometrical concentration factor is 50x. This means that in most cases the focus line is thinner than 3 cm. The performance of the lens with respect to the shape of the focal area and the position of the focal line has been analyzed with ray tracing techniques. From this analyses and by the development of a smart tracking system only two motors can bring the receivers in the required positions. One motor controls the distance between lens and receiver and the other controls the translocation of the receivers parallel to the lens. The second conclusion was that the positions of the focal line are within the bounds of the greenhouse construction for almost the whole year. Only in winter, in the early morning and at the end of the day, the focal line will be unreachable. The 480 m(2) greenhouse, with the LCPVT system based on Static Fresnel lenses and a 40 m CPVT-module and a 200 m CT-module, is designed by Bode Project Engineering and constructed by Technokas in Bleiswijk the Netherlands.

Het verbruik van aardgas door de Nederlandse glastuinbouw bedraagt ca. 10% van het totale Nederlandse verbruik. De glastuinbouwsector heeft de ambitie dat vanaf 2020 de teelt in nieuwe kassen klimaatneutraal en economisch rendabel zal zijn. Het Productschap Tuinbouw (PT), het ministerie van Economische Zaken, Landbouw en Innovatie (EL&I) en LTO Glaskracht werken samen in het programma Kas als Energiebron om deze ambitie waar te maken. In dit programma zijn een aantal transitiepaden opgesteld om deze ambitie te kunnen verwezenlijken waaronder het transitiepad zonne-energie uit het onderwerp duurzame energiebronnen.

In warm periods the excess of incoming solar energy into a greenhouse is more than required for the growth of the crop. In particular the near infrared radiation (NIR) part of the incoming radiation is not necessarily. In a previous research project a new type of greenhouse with an integrated concentrated photovoltaic system with thermal energy output (CPVT-system) was developed. This earlier system was based on a circular covering geometry and an integrated filter for reflecting the NIR of the greenhouse. The reflected radiation was used in a solar energy system. In this feasibility study the new CPVT-system is simplified so more economically by avoiding the asymmetric greenhouse construction with bended glass and the large construction for solar tracking. All parts of the solar concentrating system will be mounted inside a standard Venlo type greenhouse. The concentrator consists of lamellae which only focus the NIR-part of the spectrum onto the CPVT–module. This module is mounted to or integrated into the ridge or gutter of the greenhouse. With this spectral selection the heat load inside the greenhouse will be reduced. The target is a maximal total annual electricity production. The optimization is done with a ray tracing model fed with actual radiation data. Two types of lamellae are compared: flat lamellae and trough shaped lamellae which focus the radiation individually. Trough shaped lamellae have the advantage of reducing the number of lamellae in combination with a high concentration ratio. This will lower the costs for the servo drive of the lamellae. The reflected NIR radiation can be focused with a geometric concentration factor of 100x. The lamellae will not only reflect 49% of the NIR radiation but also a part of the whole solar spectrum. The total effective concentration ration factor will be 23x, including the transmission losses of the greenhouse and the efficiency of the concentrator. The high geometric concentration factor will limit the shadowing effect of the cultivation area by the PV-cells with only 1%. Further optimalisation in the energy yield were performed on determination of the optimal focal length of the trough shaped lamellae. The highest annual electrical output was found for lamellae with individual optimized focal lengths. In that case the annual output for Dutch climate conditions can be over 29 kWh/m²

In order to reduce the energy losses caused by shadow lines, three options are investigated. These are: 1. the use of two types of diodes; 2. the use of an "ideal" diode based an active bypass by using MOS-FET's [4] and 3. parallel switching of a number of cells between two shadow lines. The first method can reduce the voltage losses of the diode to about 300mV when oversized Schottky diodes are used. With the second method it is possible to reduce the voltage losses further to about 60mV for FETs with a resistance of 3m Omega. This method has as disadvantage that more electronic components are required to control the FET. With the third method about 10 cells are placed in parallel in one module. In that case only one shadow line appears on each module. Series connection of these parallel modules will result in zero energy losses because no bypass diodes are needed at all. This method has as a disadvantage of very high current output of the module of up to 200A. In a model, the three methods are further analyzed with respect of power losses and costs. The effect of the degree of shadowing on the daily energy yield and hence the fill factor is simulated with a model of the PV module. These results can be used in the design of the circuit of the ideal bypass diode.

De deskundigheid en de toenemende kosten voor de inzet van de arbeid in het werkproces omtrekmeting zijn een toenemend probleem voor laanboombedrijven. Meten, merken, tellen en registreren zijn vier arbeidsintensieve handelingen in het werkproces “diktemeting laanbomen” in de laanboomsector. Doel van die diktemeting (beter omtrekmeting) is het vaststellen van de omtrek van de boom (basis voor de prijs), het bepalen van het aantal verkoopbare bomen en het voorbereiden van het logistieke proces van verkoop en aflevering.

A greenhouse with Fresnel lenses in the south facing roof and a receiver for concentrated Photovoltaic with water cooling (CPVT system) will result in electrical and thermal energy output from the solar energy excess entering a greenhouse. The PV system converts about half of the direct radiation into heat and electricity. During periods with direct radiation this will significantly reduce the heat load on the greenhouse For an optimal performance the roof elements must be asymmetric with a steep inclination at the north side (the exact angle of course depends on the latitude of the building site). The Fresnel lens structure is best oriented in upwards direction. In the current design, two lenses are placed in the inner space of a double glass. This prevents pollution and condensation on the lenses. By the upward facing of the lens structure, the focus quality is preserved over a much broader range of angles of incidence compared to a lens with downward facing structures. Each PMMA lens with a size of 1.20m x 1.52m is composed of 12 ‘tiles’ for easy production. The focal distance of the lens is 1,875m and the geometrical concentration factor is 50×. This means that in most cases the focus line is thinner than 3 cm. The performance of the lens with respect to the shape of the focal area and the position of the focal line has been analyzed with ray tracing techniques. From this analyses and by the development of a smart tracking system only two motors can bring the receivers in the required positions. One motor controls the distance between lens and receiver and the other controls the translocation of the receivers parallel to the lens. The second conclusion was that the positions of the focal line are within the bounds of the greenhouse construction for almost the whole year. Only in winter, in the early morning and at the end of the day, the focal line will be unreachable. The light sum is very stable in the greenhouse compared with the light sum outside. The 480 m2 greenhouse, with the LCPVT system based on Static Fresnel lenses and a 12 m CPVT-module and a 200 m CT-module, is designed by Bode Project Engineering and constructed by Technokas in Bleiswijk the Netherlands. An electrical power of 37W/(m2 greenhouse) is measured at an incoming global radiation of 870 W/m² (on a horizontal plane). The fraction collected thermal yield is about 20% of the total incident direct radiation.