Serie Documentos
New ways to determine quality Integration of phenomics, data gathering, robotics and automation in retail
Miguel Merino-Pacheco Written by Dr. Miguel Merino Pacheco Based on the presentation at the ‘Amsterdam Produce Show’ by Rick van de Zedde, Wageningen Food and Biobased Research
Biblioteca Horticultura http://www.bibliotecahorticultura.com/
Serie Documentos
Methods to measure quality and deterioration of fresh products
New ways to determine quality Integration of phenomics, data gathering, robotics and automation in retail Written by Dr. Miguel Merino Pacheco Based on the presentation at the ‘Amsterdam Produce Show’ by Rick van de Zedde, Wageningen Food and Biobased Research
Index Introduction…………………………………………………………………. Pg. 3
Agro food robotics
Pg. 3
Quality phenomics
Pg. 4
Packaging and robotics
Pg. 5
Automation in retail
Pg. 8
Public – Private
Pg. 9
Literature Cited
Pg. 10
2 Biblioteca Horticultura http://www.bibliotecahorticultura.com/
Serie Documentos
Methods to measure quality and deterioration of fresh products
Introduction The key goal of the research project named “Quality Phenomics”, of the University of Wageningen, The Netherlands, is to develop new methods to measure quality and predict quality deterioration of vegetable and fruit products. As Rick van de Zedde1 explains, the project is a close collaboration between three different fields of expertise within Wageningen UR and several plant-breeding companies, growers, fresh-produce processors and technology providers. The combination of:
Agro Food Robotics
Quality phenomics
Robotics in packaging
Automation in retail
makes possible to keep the highest possible quality along the way from the fields to the consumer´s table.
Agro food robotics Robotics deals with the design, construction, operation, and use of robots, as well as computer systems for their control, sensory feedback, and information processing. It is an interdisciplinary branch of engineering and science that includes mechanical and electrical engineering, computer science and other related disciplines. The video “Agro Food Robotics”, explains more about.
Mr. Van de Zedde explains that agrofood is the new growth market for robotics worldwide. Even when agriculture and robotics, at first sight, do not seem to be an appropriate field of development, it is exactly the opposite. While robots are common practice in the automotive industry, the agro-food sector still uses lots of manual labor. An important reason for this is that food products come in a wide variety: No two apples are exactly the same. This makes it difficult for robots to perform sensory analyses of the products in order to replace the manual labor in agro-food processes.
Very often, robots are developed and then start the search of a convenient use for the machines. At the University of Wageningen, the angle is exactly the opposite: the research and engineering process starts with the need for an innovation. This way leads to a solution that requires the use of robotics. The special challenge for robotics in agriculture is that it is necessary to deal with a large natural variation. 1
Senior researcher/ business developer for Computer Vision at Wageningen Food & Biobased Research http://www.wur.nl/en/Persons/drs.-HJ-Rick-van-de-Zedde.htm
3 Biblioteca Horticultura http://www.bibliotecahorticultura.com/
Serie Documentos
Methods to measure quality and deterioration of fresh products
Robotics allows, among other solutions, to collect data on an unprecedented scale. Regarding the search of quality, many quality criteria cannot be observed by the human eye during the production process, and manual work is both expensive and unreliable. New technology developments, however, make it possible to quickly carry out objective quality inspections of food products from farm to table.
Robots furnished with special cameras, optical sensors and computer vision systems can do the job. In the presentation advanced approaches for both two- and three-dimensional visual inspection, objective color measurement, and near infrared (NIR) hyperspectral inspection that reveals “invisible” deviations, along with advanced methods for automatic image processing are explained. The massive collection of information using non-destructive methods allows to narrow radically the confidence intervals in the values of the variables defining produce quality.
Quality phenomics Phenomics is an area of biology concerned with the measurement of phenomes—the physical and biochemical traits of organisms as they change in response to genetic mutation and environmental influences. We have in the one hand a genome – a genetic basis – and in the other different expressions – phenotypes – of that genome when exposed to different environments. Quantitative trait loci (QTL) and genes associated to quality traits have been identified in vegetables. And many of them show environment interaction effects. “To gain significant progress in improving the quality and productivity of crops, it is necessary to understand and exploit the genotype and the phenotype of plants/ products in a changing environment (climate change/ diseases)”. The ripening of subtropical fruit – mango, avocado – represents a good example of this situation. These fruits show a huge variability in quality, both on the fields and also on the supermarket shelves. And the produce - which of course presents also genetic variability - is subjected also to a wide array of conditions during cropping, transportation and storage. The question is then, how produce can ripe in a more accurate and efficient form? The proposed way is to measure the initial ripeness on individual fruit pieces, understand the development of quality traits along time and apply then customized ripening protocols.
4 Biblioteca Horticultura http://www.bibliotecahorticultura.com/
Serie Documentos
Methods to measure quality and deterioration of fresh products
Identifying by the way the reaction of the genetic load to all that environmental variation and its manifestation – quality. Sensors and automation allow to obtain the huge amount of data needed for that process. Final quality – involving appearance, organoleptic and nutritional aspects - is predicted through establishing initial quality for concrete genomes, the conditions of the chain and the development of the quality – reaction of the genomes - under such conditions.
The measurement of different quality traits can be automated. Firmness is a very important quality characteristic in fruit, and its evolution until the final stage – ready to eat – can be studied through non-destructive acoustic measurement. In avocado, acoustic firmness measurement has a strong correlation with fruit texture analysis, though location of the kernel influences the signal. And in mango, measuring firmness with limited compression (penetrometer) has a high correlation (r2 around 0.9) with acoustic stiffness measurement (ticking a hammer to the fruit and listening to the sound). In pears, percentage of brown spots and color could be established using sensors. One aim is to reach on-crate level inspection in order to achieve sub-sampling of products. A job that presently is done by manual workers. Non-destructive sensors placed on a robot arm allows to perform that job in automated form. The gap between lab and current daily practice is being presently bridged. Brix measurement in tomato using a self– made NIR (near infrared) probe and mounted in a robot arm for automation have been performed.
Packaging and robotics The fruit and vegetable packaging industry faces the challenge of being flexible. Large number of products and packaging variations, small batches, late placed orders… all of that place hard demands that are met, up to now, by manual labor, because people are flexible.
The project PicknPack (http://www.picknpack.eu) is developing a system to make possible to pack several different products on the same production line. That means, through automation, robotics and information integration obtain the same flexibility of a human managed line, increasing accuracy and productivity at the same time.
5 Biblioteca Horticultura http://www.bibliotecahorticultura.com/
Serie Documentos
Methods to measure quality and deterioration of fresh products
It is a flexible robotic system which will develop modules that can cope with the typical variability of food products and the requirements of the sector regarding hygiene, economics and adaptability. It consists of three types of modules: a) for food quality inspection – to asses quality of produce before or after packaging; b) for robotic food handling – a vision controlled handling module that picks up and separates produce from a harvest bin and places it in a package; and c) for adaptive packaging to accommodate various types of packaging with flexibility for size, shape, product environment, sealing and printing.
More specifically, those functions are developed, in several smart modules, as follows: Module 1) Flexible thermoformer. The package is shaped to the exact size and shape of the product, fitting practically any kind of fruit. Module 2) As the plastic is molded, a RFID scanner examines the crop crate and establishes when and where the product was harvested. The information goes to the line controller. Module 3) Pick and place robot. This robot collects product from the harvest crate and places it in moulds. It uses a vision controlled robotic gripper. Which is automatically cleaned in the robot cell, when a different batch of produce is on the line. Module 4) After picking the product, it is time for a quality check. It is done by a Quality Assessment System (QAS), consisting in a variety of sensors. This QAS-module comprises of five separate vision systems that together can measure various properties over a broad range of the electromagnetic spectrum. The module comprises two separate submodules. Submodule A houses an X-ray vision system (InnoS) and Submodule B houses a hyperspectral camera (KUL), an RGB camera (DLO), a 3D imaging unit (DLO) and a microwave sensor (UM). The cameras are mounted above the transporting system and inspect the content of all packages that pass below.
Quality Assessment System QAS (Source: Project PicknPack)
6 Biblioteca Horticultura http://www.bibliotecahorticultura.com/
Serie Documentos
Methods to measure quality and deterioration of fresh products
Module 5) The surface of the package is personalized printed with a flexible printer. There is no need for labels or pre – printed foils. The exterior and / or final destination can be determined at the very last moment. A laser sealer cuts the packages into any desired shape and seals it. A package audit with an infrared camera takes place and resealing is performed, when necessary. Module 6) Each package is provided with a RFID tag, a small antenna that allows to track down the product after it is checked, printed and sealed. Module 7) The packages are sorted according to quality and destination. The innovative sorting system called ―Pickable‖ hardly uses any floor space, despite its large and flexible range.
The entire surface of the system never has to be cleaned by hand. A self-driving cleaning robot cleans each module. Those modules can be used independently, in all sorts of different configurations.
The smart modules of PicknPack, assembled to a standard configuration
7 Biblioteca Horticultura http://www.bibliotecahorticultura.com/
Serie Documentos
Methods to measure quality and deterioration of fresh products
Automation in retail In retail outlets, fully automated solutions are already available. The picture shows the EasyFlow automated checkout ITAB (http://www.itab.se/lav/ITAB-Scanflow/Home/Products/EasyFlow/)
EasyFlow automated check out The fully automated checkout uses six different sensor technologies – computer vision, near infrared technology, weight and statistics - and nine different classifiers linked to a unique network. It yields a product identification ratio of more than 99% - all without the need of barcodes. The system improves its functioning through continuous self-learning. The only thing the customer needs to do is to put the articles on the belt. The rest is handled by the system and the articles appear on the receipt.
8 Biblioteca Horticultura http://www.bibliotecahorticultura.com/
Serie Documentos
Methods to measure quality and deterioration of fresh products
Public – Private The public – private partnerships arranged to develop these kind of research combine the knowledge push of the research partners with the market pull of the market knowledgeable participants. Fundamental research is provided by universities – like Wageningen - , oriented with applied research following implementation needs provided by the markets. Sensor and robot technology for non-destructive quality inspection of fruit and vegetables and automatic collection of data is being further developed by the Wageningen UR Food & Biobased Research. Genetics and environmental change are measured and correlated thanks to the large data collecting capacity of robotics, and that knowledge applied to make first quality produce reach the consumer.
9 Biblioteca Horticultura http://www.bibliotecahorticultura.com/
Serie Documentos
Methods to measure quality and deterioration of fresh products
Literature Cited ZEDDE, Rick van de. (2017). New ways to determine quality Integration of phenomics, data gathering, robotics and automation in retail. 31 January 2017, of the web Biblioteca Horticultura: http://publicaciones.poscosecha.com/es/ (*) Written by Dr. Miguel Merino Pacheco, Mittelmeer Unternehmenskommunikation, Eschwege – Deutschland, mittelmeer@gmail.com, based on the presentation at the Amsterdam Produce Show, November 2016 by Rick van de Zedde, Wageningen Food and Biobased Research.
Dr. Manuel Candela, 26 11ª - 46021 Valencia, España Tel +34 – 649 485 677 info@poscosecha.com info@bibliotecahorticultura.com http://www.poscosecha.com
http://www.postharvest.biz
http://www.horticulturablog.com http://www.bibliotecahorticultura.com
http://www.actualfruveg.com
10 Biblioteca Horticultura http://www.bibliotecahorticultura.com/