Smart Sensors from Ground to Cloud and Web Intelligence

Proceedings, 6th IFAC Conference on Bio-Robotics Proceedings, IFAC Conference on Bio-Robotics Beijing, China,6th July 13-15, 2018 Proceedings, IFAC Conference on Bio-Robotics Beijing, China,6th July 13-15, 2018 online at www.sciencedirect.com Proceedings, 6th IFAC Conference onAvailable Bio-Robotics Proceedings, IFAC Conference on Bio-Robotics Beijing, China, China,6th July 13-15, 2018 Beijing, July 13-15, 2018 Beijing, China, July 13-15, 2018

ScienceDirect

IFAC PapersOnLine 51-17 (2018) 31–38

Smart Sensors from Ground to Cloud and Web Intelligence Smart Sensors from Ground to Cloud and Web Intelligence Smart Sensors from Ground to Cloud and Web Intelligence Smart from Ground Cloud and Intelligence Wu Yong*, Li Shuaishuai*, Li Li*, Li Minzan**, Li Ming***, Smart Sensors Sensors from Ground to to Cloud and Web Web Intelligence

Wu Yong*, Li Shuaishuai*, Li Li*, Li Minzan**, Li Ming***, K.G. Arvanitis****, Cs. Georgieva*****, N. Sigrimis**** Wu Yong*, Li Shuaishuai*, Shuaishuai*, Li Li*, Li*, Li Li Wu Yong*, Li Li Li Minzan**, Minzan**, Li Ming***, Ming***, K.G. Arvanitis****, Cs. Georgieva*****, N. Sigrimis**** Wu Yong*, Li Shuaishuai*, Li Li*, Li Minzan**, Li Ming***, K.G. N. Sigrimis**** Sigrimis**** K.G. Arvanitis****, Arvanitis****, Cs. Cs. Georgieva*****, Georgieva*****, N. K.G. Arvanitis****, Cs. Georgieva*****, N. Sigrimis**** *Key Laboratory of Modern Precision Agriculture System Integration Research, Ministry of Education, *Key Laboratory of Modern Precision Agriculture System Integration Research, Ministry of Education, China Agricultural University, Beijing 100083 China (e-mail: [email protected]) *Key Laboratory Laboratory of Modern Modern PrecisionBeijing Agriculture System Integration Research, Ministry Ministry of of Education, Education, *Key of Precision Agriculture System Integration Research, China Agricultural University, 100083 China (e-mail: [email protected]) Laboratory of Agricultural Information Acquisition Technology, Ministry of Agriculture, *Key**Key Laboratory of Modern PrecisionBeijing Agriculture System Integration Research, Ministry of Education, China Agricultural University, 100083 China (e-mail: [email protected]) China Agricultural University, Beijing 100083 China (e-mail: [email protected]) **Key Laboratory of Agricultural Information Acquisition Technology, Ministry of Agriculture, China Agricultural University, Beijing 100083, (e-mail: [email protected]) China Agricultural University, Beijing 100083 ChinaChina (e-mail: [email protected]) **Key Laboratory of Information Acquisition Technology, Ministry of **Key Laboratory of Agricultural Agricultural Information Acquisition Technology, Ministry of Agriculture, Agriculture, China Agricultural University, Beijing 100083, China (e-mail: [email protected]) ***Beijing Research Center for Information Technology in agriculture, Beijing 100097, **Key Laboratory of Agricultural Information Acquisition Technology, Ministry of Agriculture, China Agricultural University, Beijing 100083, China (e-mail: [email protected]) China Agricultural University, Beijing 100083, China (e-mail: [email protected]) ***Beijing Research Center for Information Technology in agriculture, Beijing 100097, China (e-mail: [email protected]) China Agricultural University, Beijing 100083, China (e-mail: [email protected]) ***Beijing Research Center for Information Technology in Beijing 100097, ***Beijing ResearchResources Center forManagement Information Technology in agriculture, agriculture, 100097, China (e-mail: [email protected]) ****Department of Natural & Agricultural Engineering,Beijing Agricultural University ***Beijing Research Center for Information Technology in agriculture, Beijing 100097, China (e-mail: [email protected]) China (e-mail: [email protected]) ****Department of Natural Resources Management & Agricultural Engineering, Agricultural University of Athens, Athens 11855, Greece (e-mail: [email protected]) China (e-mail: [email protected]) ****Department of of Natural Natural Resources Management & Agricultural Agricultural Engineering, Agricultural Agricultural University University ****Department Resources Management & Engineering, ofAutomatics Athens, Athens Greece (e-mail: [email protected]) *****Department of and 11855, Mechatronics, University of Ruse, RuseAgricultural 7017, Bulgaria ****Department of Natural Resources Management & Agricultural Engineering, University of Athens, Athens 11855, Greece (e-mail: [email protected]) of Athens, Athens 11855, Greece (e-mail: [email protected]) *****Department of Automatics and Mechatronics, University of Ruse, Ruse 7017, Bulgaria of Athens, Athens 11855, Greece (e-mail: [email protected]) *****Department of of Automatics Automatics and and Mechatronics, Mechatronics, University University of of Ruse, Ruse, Ruse Ruse 7017, 7017, Bulgaria Bulgaria *****Department *****Department of Automatics and Mechatronics, University of Ruse, Ruse 7017, Bulgaria Abstract: With the development of sensing, wireless communication, and Internet technologies, we are Abstract: the development sensing, wireless and Internet technologies, are now livingWith in a world that is filledof various smartcommunication, things – the Internet of Things (IOT). Thiswe paper Abstract: With the development development ofwith sensing, wireless communication, and Internet Internet technologies, we are Abstract: the sensing, wireless communication, and technologies, we are now livingWith in aprospects world thatanisemerging filledof with various smart things – the Internet of Things (IOT). Thisaims paper introduces and research area – the Embedded Intelligence (EI). This field at Abstract: With the development ofwith sensing, wireless communication, and Internet technologies, we are now living in aaprospects world that is filled various smart things – the Internet of Things (IOT). This paper now living in world that is filled with various smart things – the Internet of Things (IOT). This paper introduces and an emerging research area – the Embedded Intelligence (EI). This field aims at revealing the individual behaviours, spatial contexts, as well as social patterns and urban dynamics by now living and in aprospects world thatanisemerging filled with variousarea smart things – the Internet of Things (IOT). Thisaims paper introduces research – the Embedded Intelligence (EI). This field at introduces and prospects an emerging research area – the Embedded Intelligence (EI). This field aims at revealing the individual behaviours, spatial contexts, as well as social patterns and urban dynamics by mining thethe digital traces behaviours, leftemerging by people while interacting withasInternet of smartand things (cameras, smart introduces and prospects an research area – the Embedded Intelligence (EI).urban This field aims at revealing individual spatial contexts, as well social patterns dynamics by revealing the individual behaviours, spatial contexts, as well as social patterns and urban dynamics by mining the digital traces left by people while interacting with Internet of smart things (cameras, smart cars, smart cards, etc). Inbehaviours, the by agricultural sector we add mining of existing technology (books, articles, revealing the individual spatial contexts, as well asInternet social patterns and urban dynamics by mining the digital traces left people while interacting with of smart things (cameras, smart mining the cards, digital tracesIn left by people while interacting with of smart things (cameras, smart cars, prints) smart etc). thedepth agricultural sector we add mining of existing technology (books, articles, blue to generate high ontologies or metadata and Internet knowledge extraction from Big Data for mining the cards, digital tracesIn left by people while interacting with Internet of smart things (cameras, smart cars, smart etc). the agricultural sector we add mining of existing technology (books, articles, cars, smart cards, etc). In the agricultural sector we add mining of existing technology (books, articles, blue prints) to generate high depth ontologies or metadata and knowledge extraction from Big Data for reasoning and making pro-active decisions using open data information. We include intelligent data cars, smart cards, etc). In the agricultural sector we add mining of existing technology (books, articles, blue prints)and to generate generate high depth depth decisions ontologiesusing or metadata metadata andinformation. knowledge extraction extraction from Big Data Datadata for blue prints) to high ontologies or and knowledge from Big for reasoning making pro-active open data We include intelligent analysis toand discover new from using data records. The paper discusses thefrom research history, blue prints) to generate highknowledge depth decisions ontologies or metadata and knowledge extraction Big Data for reasoning making pro-active open data information. We include intelligent data reasoning and making pro-active decisions using open data information. We include intelligent data analysis to discover new knowledgemajor fromapplications, data records. The paper issues discusses theandresearch history, characteristics, general architecture, and research of EI exemplifies an reasoning and making pro-active decisions using open data information. We include intelligent data analysis to to discover discover new knowledgemajor fromapplications, data records. records. The paper issues discusses theandresearch research history, analysis new knowledge from data paper discusses the history, characteristics, general architecture, andThe research ofaEI exemplifies an application of the team in IOT smart irrigation. The purpose of presenting it as Technology Roadmap analysis to discover new knowledgemajor fromapplications, data records. The paper issues discusses theandresearch history, characteristics, general architecture, and research of EI exemplifies an characteristics, general architecture, major applications, and research issues of EI and exemplifies an application of the team in IOT smart irrigation. The purpose of presenting it as a Technology Roadmap (TRM) wasoftothe clarify the challenges and applications, opportunities in of the general area “Intelligent Building characteristics, general architecture, major and research issues ofof EI and exemplifies an application team in IOT smart irrigation. The purpose presenting it as a Technology Roadmap application of the team in IOT smart irrigation. The purpose of presenting it as a Technology Roadmap (TRM) was to clarify the challenges and opportunities in the general area of “Intelligent Building Technologies” (IBT) more specifically toThe smart greenhouses, smart irrigation, or smart crop application oftothe teamand in IOT smart irrigation. purpose of presenting it as a Technology Roadmap (TRM) was clarify the challenges and opportunities in the general area of “Intelligent Building (TRM) was toThe clarify the more challenges and agriculture opportunities in the general of “Intelligent Building Technologies” (IBT) and specifically to smart isgreenhouses, irrigation, crop management. understanding of smart important tosmart bearea identified byor thesmart scientific (TRM) was to (IBT) clarify the more challenges and opportunities in the general area of “Intelligent Building Technologies” and specifically to smart greenhouses, smart irrigation, or smart crop Technologies” (IBT) and more specifically to smart greenhouses, smart irrigation, or smart crop management. The understanding of smart agriculture is important to be identified by the scientific community as The a(IBT) significant issue because it agriculture is to needed nations to develop and adoptbythese emerging Technologies” and more specifically smartfor greenhouses, smart irrigation, orthesmart crop management. understanding of smart is important to be identified scientific management. The understanding of smart agriculture is important to be identified by the scientific community as a significant issue because it is needed for nations to develop and adopt these emerging technologies. Global interoperability is of key importance, and so it is a common understanding of values management. The understanding of smart agriculture is important to be identified by the scientific community asGlobal significant issue because because it is isimportance, needed for for and nations to develop andunderstanding adopt these these emerging emerging community as aaand significant issue it needed nations and adopt technologies. interoperability of key sointernational it to is develop a common of values such as privacy security, basedbecause onis open, and transparent standards. community asGlobal a significant issue it fair isimportance, needed for and nations to develop andunderstanding adopt these emerging technologies. interoperability is of key so it is a common of values values technologies. Global interoperability is of key importance, and so it is a common understanding of such as privacy and security, based on open, fair and transparent international standards. technologies. Global interoperability ofAutomatic keyfair importance, and sointernational it isbya Elsevier common understanding of values such as privacy privacy and of security, based onisopen, open, andControl) transparent standards. © 2018, IFAC (International of Hosting Ltd. All rights reserved. such as and security, based on fair and transparent international standards. Keywords: internet things,Federation ubiquitous computing, embedded intelligence, precision irrigation. such as privacy and of security, on open, fair and embedded transparentintelligence, internationalprecision standards. Keywords: internet things, based ubiquitous computing, irrigation. Keywords: internet of of things, things, ubiquitous ubiquitous computing, computing, embedded embedded intelligence, intelligence, precision precision irrigation. irrigation. Keywords: internet Keywords: internet of things, ubiquitous computing, embedded intelligence, precision irrigation. process that is designed to help industry, its supply-chain, 1. INTRODUCTION process thatand is designed help industry, its supply-chain, academic research to groups, and governments come 1. INTRODUCTION process that is designed to help industry, its supply-chain, process that is designed to help industry, its supply-chain, academic and research groups, and governments come 1. INTRODUCTION together to jointly identify and prioritize the technologies Agriculture will be 1.renewed through digitalisation, and its process INTRODUCTION that is designed to help industry, its supply-chain, academictoand and research groups, and governments governments come academic research groups, and come INTRODUCTION jointly identify and prioritize the technologies Agriculture will will be 1.renewed through Such digitalisation, and its needed totosupport strategic research (R&D), competitiveness be reinforced. a revolution is together academic and research groups, and development governments come together jointly identify and prioritize the technologies Agriculture will be renewed through digitalisation, and its together to jointly identify and prioritize the technologies Agriculture will be renewed through digitalisation, and its needed to support strategic research and development (R&D), competitiveness will be reinforced. Such a revolution is marketing and investment decisions. These technologies will already happening but in larger farms and facilities. Drones, together to jointly identify and prioritize the technologies Agriculture will be renewed through digitalisation, and its needed to support strategic research and development (R&D), competitiveness will be reinforced. Such revolution is marketing needed to support strategic research and development competitiveness will reinforced. Such aa sensors, revolution is andimportance investment These already happening but be in larger farms and facilities. Drones, of critical todecisions. an industry in thetechnologies next five(R&D), towill ten milking robots, unmanned vehicles, intelligent dataneeded to support strategic research and development (R&D), competitiveness will be reinforced. Such a revolution is be marketing and investment decisions. These technologies will already happening but in larger farms and facilities. Drones, marketing and investment decisions. These technologies will already happening but in larger farms and facilities. Drones, be of critical importance to an industry in the next five to ten milking robots, unmanned vehicles, intelligent sensors, datayears that blockchain will be dominating advances. In sharing etc., all is calling farmers to embrace a “digital marketing andimportance investmenttodecisions. These technologies will already happening but in larger farms and facilities. Drones, be of an in next five to milking robots, vehicles, intelligent sensors, databe of critical critical importance to anisindustry industry in the the next five to ten ten milking robots, unmanned vehicles, intelligent sensors, dataIn years that we blockchain will bepresently dominating advances. sharing etc., However, allunmanned is calling farmers to embrace a “digital agriculture could state it needed to face the revolution”. financial investment necessary for be of critical importance to an industry in the next five to ten milking robots, unmanned vehicles, intelligent sensors, data- years that blockchain will be dominating advances. In sharing etc., all is calling farmers to embrace aa “digital years that blockchain will be dominating advances. In sharing etc., all is calling farmers to embrace “digital agriculture we could state it is presently needed to face the revolution”. However, financial investment necessary for hunger of we people and the global competition. Chains of establishing isfinancial rather unrealistic for necessary mosta cases in years that blockchain will bepresently dominating advances. In sharing etc.,agriculture all is calling farmers to embrace “digital agriculture could state it is needed to face the revolution”. However, investment for agriculture we could state it is presently needed to face the revolution”. However, financial investment necessary for hunger of people and the global competition. Chains of establishing agriculture is rather unrealistic for most cases in manufacturers, transportation businesses, logistics firms, food the current context. Farms must increase production of food agriculture we could state it is presently needed to face the revolution”. However, financial investment necessary for hunger of and the competition. Chains of establishing agriculture is for cases in hunger of people people and and the global global competition. Chains of establishing agriculture is rather rather unrealistic for most most cases in transportation businesses, logistics firms, food the current context. Farms must unrealistic increase of food and energy companies, enterprises across Europe while preserving the environment, but production climate change is manufacturers, hunger of people and the other global competition. Chains of establishing agriculture is rather unrealistic for most cases in manufacturers, transportation businesses, logistics firms, food the current context. Farms must increase production of food manufacturers, transportation businesses, logistics firms, food the current context. Farms must increase production of food and energy companies, and other enterprises across Europe while preserving the environment, but climate change is are employing radio frequency identification (RFID) to cut already negatively impacting agricultural production manufacturers, transportation businesses, logistics firms, food the current context. Farms must increase production of food and employing energy companies, companies, and other other enterprises (RFID) across Europe Europe while preserving preserving the environment, environment, but climate climate production change is is and energy and enterprises across while the change are radio frequency identification torates, cut already negatively impacting enhance visibility, asset-utilization globally. Climate-smart agriculture, agricultural asbut defined andproduction presented and energy companies, and improve other enterprises across Europe while the environment, but climate change is costs, are employing radio frequency identification (RFID) to cut alreadypreserving negatively impacting agricultural are employing radio frequency identification (RFID) to cut already negatively impacting agricultural production costs, enhance visibility, improve asset-utilization rates, globally. Climate-smart agriculture, as defined and presented streamline business processes, improve inventory accuracy by FAO in 2010, is an impacting approach to developing theproduction technical, are employing radio frequency identification (RFID) torates, cut already negatively agricultural costs, enhance visibility, improve asset-utilization globally. Climate-smart agriculture, as defined and presented costs, enhance visibility, improve asset-utilization rates, globally. Climate-smart agriculture, as defined and presented streamline business processes, improve inventory accuracy by FAO in 2010, is an approach to developing the technical, achieve manyvisibility, other benefits. policy and investment conditions achieveand sustainable costs, enhance improve asset-utilization rates, globally. Climate-smart agriculture, astodefined presented and streamline business processes, improve inventory accuracy by FAO in 2010, is an approach to developing the technical, streamline processes, by FAOand in 2010, is an approach to developing thesustainable technical, achievebusiness many other benefits.improve inventory accuracy policy investment conditions to achieve agricultural development for food security under climate and streamline business processes, improve–IOTinventory by FAOand in 2010, is an approach to developing thesustainable technical, and achieve many benefits. policy investment conditions to achieve Historically, the other Internet of Things is a accuracy concept and achieve many other benefits. policy and investment conditions to achieve sustainable agricultural development for food security under climate change. Research must develop the technical solutions and achieve many other benefits. policy and investment conditions to achieve sustainable Historically, the Internet of Things –IOTis a concept agricultural development for food security under climate wherein objects are uniquely identified by barcodes, RFID agricultural development for food security under climate change. Research must develop the technical solutions Historically, the Internet of Things –IOTis a concept expected thedevelopment IOT to bring farming but under we must also Historically, the are Internet of object Things –IOTis a readings concept wherein objects uniquely identified by barcodes, RFID agricultural fortofood security climate change. Research must develop the technical solutions tags, etc. During the physical lifecycle, event change. Research must develop the aspects technical solutions the are Internet of identified Things –IOTis a concept expected the IOT to bring to farming but weandmust also Historically, wherein objects uniquely by barcodes, RFID embrace the policy and investment provide wherein objects are uniquely identified by barcodes, RFID change. Research must develop the technical solutions tags, etc. During the physical object lifecycle, event readings expected the the IOT IOT to and bringinvestment to farming farmingaspects but we weandmust must also are made using are sensors (RFID & lifecycle, NFCbyreaders, cameras, expected to bring to but also wherein objects uniquely identified barcodes, RFID embrace policy provide tags, etc. During the physical object event readings solutions ODT (One Dollar a Thing, Sigrimis 2015). The tags, etc. During the physical object lifecycle, event readings expected the IOT to bring to farming but we must also are made using sensors (RFID & NFC readers, cameras, embrace the the policy and investment aspects and provide scanners, GPS/GSM/Wi-Fi, manual reading, etc.), collected embrace policy and investment aspects and provide tags, etc. During the physical object lifecycle, event readings solutions ODT (One Dollar a Thing, Sigrimis 2015). The are made using sensors (RFID & NFC readers, cameras, Technology Roadmap (TRM) concept is a and consultative using sensors (RFID & orreading, NFC readers, cameras, scanners, GPS/GSM/Wi-Fi, manual etc.), embrace policy investment aspects provide solutions the ODT (One and Dollar Thing, Sigrimis 2015). The are and made recorded in databases, private on the cloud.collected Further solutions ODT (One Dollar aa Thing, 2015). The are made using sensors (RFID & NFC readers, cameras, Technology Roadmap (TRM) conceptSigrimis is a consultative GPS/GSM/Wi-Fi, manual scanners, GPS/GSM/Wi-Fi, manualorreading, reading, etc.), collected and recorded in databases, private on the etc.), cloud.collected Further solutions ODT (One Dollar a Thing, Sigrimis 2015). The scanners, Technology Roadmap (TRM) concept is a consultative Technology Roadmap (TRM) concept is a consultative scanners, GPS/GSM/Wi-Fi, manual reading, etc.), collected and recorded recorded in in databases, databases, private private or or on on the the cloud. cloud. Further Further and Technology Roadmap (TRM) concept is a consultative Copyright 2018 IFAC 31 Hosting 2405-8963 © 2018, IFAC (International Federation of Automatic Control) by Elsevier Ltd. All rights reserved. and recorded in databases, private or on the cloud. Further Copyright 2018 responsibility IFAC 31 Control. Peer review©under of International Federation of Automatic Copyright © 2018 2018 IFAC IFAC 31 10.1016/j.ifacol.2018.08.057 Copyright © 31 Copyright © 2018 IFAC 31

IFAC BIOROBOTICS 2018 32 Beijing, China, July 13-15, 2018

Wu Yong et al. / IFAC PapersOnLine 51-17 (2018) 31–38

these records are aggregated, consolidated or merged with other information already handled in computer systems for traceability, logistics, management or sales issues. The original idea of the Auto-ID Center is based on RFID-tags and unique identification through the Electronic Product Code (EPC). The EPC is designed as a unique universal identifier for every physical object anywhere in the world, for all time. Its structure is defined in the EPCglobal Tag Data Standard, which is an open standard freely available for download from the website of EPCglobal, Inc. The next generation of Internet applications using Internet Protocol Version 6 (IPv6) would be able to communicate with devices attached to virtually all human-made objects because of the extremely large address space of the IPv6 protocol.

purpose, and is synonymous to pervasive communications or to the SED (=Speaking Electronic Devices), as we named such IOT predecessor in the 90s. It is also referring to the exploitation of the data that is of course a necessity for adding value to IOT, but it is not necessary for the IOT characterization. Of course, the fact is that the definition probably will need frequent updates to reflect the evolution of the state of the art, but the “simplest the more lasting”, until the technology time has matured enough to need no definitions or end the debate. The research area that was opened with IOT is enormous, ranging from Technologies and concepts in modern networked things, to Emerging IOT business models and process changes. Further onto the “Smart Objects” that is supported by Embedded Intelligence, a never-ending subject as it will need to catch human intelligence, which has no time limit ever. The Industrial Internet of Things (IIoT) is the use of Internet of Things (IoT) technologies in manufacturing. Also known as the Industrial Internet, IIoT incorporates machine learning and big data technology, harnessing the sensor data, machine-to-machine (M2M) communication and advancing automation technologies that have existed in industrial settings for years. A major concern, surrounding the Industrial IoT and AgIOT, is interoperability between devices and machines that use different protocols and have different architectures. The nonprofit “Industrial Internet Consortium”, founded in 2014, focuses on creating standards that promote open interoperability and the development of common architectures.

A combination of these ideas can be found in the current GS1/EPC global EPC Information Services (EPCIS) specifications. This system is being used to identify objects in industries ranging from Aerospace to Fast Moving Consumer Products and Transportation Logistics. Can we define the Internet of Things in a way that is concise, clear, general, but specific enough to be comprehensive? The question of the precise definition of the IOT has created considerable intellectual debate in recent years. Getting the definition perfect is important for regulatory, legal, legislative and scientific purposes. Several presentations at the European Commission (Uckelmann et al 2011) and exchanges on a mailing list continue the debate yet. The definition proposed by Monique Morrow of Cisco, suggests (1): "The Internet of Things consists of networks of sensors attached to objects and communications devices, providing data that can be analyzed and used to initiate automated actions. The data also generates vital intelligence for planning, management, policy and decision-making."

In the 80’s it was not possible to contribute much in the field of AI applications for agriculture because AI was not mature yet in its generic field. Today it can be declared that we need a technology, like IOT, as the enabler for exploitation of knowledge at field level, or as said, “from the research lab to the root of the plant”-Geomations Logo. In ecosystems natural creations such as streams, rivers, coastal wetlands, grasslands, and forests provide numerous services that fundamentally support human health and well-being. These also need protection from climate change and human activity likewise the production agriculture.

Olivier Dubuisson of Orange FT Group defines Internet of Things as: "A global ICT infrastructure linking physical objects and virtual objects (as the informational counterparts of physical objects) through the exploitation of sensor & actuator data capture, processing and transmission capabilities. As such, the IOT is an overlay above the 'generic' Internet, offering federated physical-object-related services (including, if relevant, identification, monitoring and control of these objects) to all kinds of applications".

The importance of electronics in the modern world is hard to overstate, touching every aspect of life. A good example is the evolution of telephone followed by Tablets that offer productivity tools for enterprises, healthcare institutions and governments, enhancing communications and providing a mobile communications work platform. Telecommunications offers a striking example of the rapidity of the electronics revolution. The move from 1G to 4G took a full decade but the time to 5G and 6G will be much shorter with more plentiful of services.

In our opinion, the IOT definition should not specify the purpose. It must describe only what it is, and in that sense we agree with Monique as it is concise, clear and comprehensive but to be a little more general (as it should) we would avoid the necessity of networks (it may be only one connected to the internet) and no sensors (it may not be a sentient environment but just an RFID), so we would state that: “The Internet of Things consists of devices attached to objects and communications devices, providing data on the Internet that can be analysed and used to feedback or initiate automated actions”. This includes the widely accepted form of a simplest RFID, the IOT generic technology (Sigrimis 2008) but also can be as big as a whole Industrial computer system connected with wires to the internet. We think the expression for “The data also generates vital intelligence for planning, management, policy and decision-making” refers to the

2. STATE OF THE ART 2.1 IPv6 With the growth of IP connected devices, the Internet is exhausting the presently still used IPv4 address protocol (of only 3.4 Billion unique addresses). Service providers anticipated (Ravinder S. 2011), that will have no longer 32

IFAC BIOROBOTICS 2018 Beijing, China, July 13-15, 2018

Wu Yong et al. / IFAC PapersOnLine 51-17 (2018) 31–38

access to additional IPv4 addresses beginning 2013. Geomations SA (our spinoff company) follows this wave to “empowered by IPv6” with a new connectivity design which will connect all different equipment to the Internet and will have contributions with cloud services (Fig. 1) for its customers but will also contribute with embedded intelligence content. We are starting operating experimental IPv6 networks in private networks and will try seamless transition from IPv4. We are in the process of IPv6_ready IOT Consumer products and prepare a full range of Agricultural Business Services (WOT) with IPv6 capabilities, so we will be ready to support IPv6 as the industry adopts it. IPv6 adoption is critical to all nations as the investments for infrastructure growth must continue with next generation capabilities. A smooth transition to the new protocol across each nation and in synch with the globe is a necessity. IDA GEO/LIOT

Kowledge Structure

A chief component of these devices is wireless communication technology. In particular, the IEEE 802.15.4 standard is very promising for the lower (physical and link) layers. As for higher layer functions, there is considerable interest from non-IETF groups in using IP technology. The IEEE 1451.5 standard for wireless transducers has a chapter for 6LoWPAN and the ISA SP100 standard for wireless industrial networks has adopted 6LoWPAN for their network layer. 2.3 Intelligent Data Analysis AI has always been concerned with algorithms and representations, but we also need to understand how to put various parts together into complete working systems, within an architecture. It is now common in AI and cognitive science to think of humans and other animals. Furthermore, many intelligent robots and software agents, have a virtual machine information processing architecture which includes different layers, and which, in the case of animals, evolved at different stages. Bio-inspired solutions today is the way to solve complex problems of life including agriculture. In Intelligent Data Analysis statistical inference and modelling are indispensable for analysing big data, such as use models to aggregate data from different sources and the very basics of Bayesian statistics for predictive modelling (Li J et al 2017).

Ontologies-CRMs CAU/AUA/FAO Evaluate CRM

Smart tools

DEEP KNOWLEDGE-MAS/Artificial Life (Plants, Insects, etc) Model identification, optimization i.e. WUM=Water Uptake Module

MAS=MultiAgent system Reasoning, Context, water allocation

SHALLOW KNOWLEDGE – expert mode Client properties, Controller mode A or B, schedules

A problem surrounding the study of learning architectures is the diverse of aims. Some researchers try to solve engineering problems by use of deep learning algorithms and care only about how well their solutions work, not paying attention on whether they model natural systems. Other researchers attempt to understand and model humans, or other animals. A few are attempting to focus only on general principles equally applicable to natural and artificial systems. For instance, a confusion arises when someone who describes a system as ‘learning’ may merely mean that it adaptively solves an engineering problem using predictive recursive analytics.

Examine expert scenarios

wDssCF

Data Input from wDBC

33

Work Schedules or URDI Parameters

URDI=Ultimate Regulated Deficit Irrigation

Fig. 1. Knowledge structure organization suggested for agricultural applications

2.2 6LoWPAN

2.4 Web semantics

Well-established fields such as control networks, and escalating ones such as "sensor" (or transducer) networks, are increasingly being based on wireless technologies. Most (but certainly not all) of these nodes are amongst the most constrained that have ever been networked wirelessly. Extreme low power (such that they will run potentially for years on batteries) and extreme low cost (total device cost in single digit dollars, and riding Moore's law to continuously reduce that price point) are seen as essential enablers towards their deployment in networks with the following characteristics:

The Semantic Web technologies could, in theory, extend the search capabilities. Today, they allow to structure in a formal way the information published on the Web as metadata. This way of structuring can give contextual sense to published information and can be handled automatically in search engines or in exchanges between applications. Reusing and adapting current Web technologies would allow anybody to automatically publish or search object-related information guaranteed to be directly usable and significant. This association, of semantic web technologies with the IoT Big Dara defines what will be the “Web 3.0”.

• Significantly more devices than current local area networks with exponential increase on routing paths and message transfer time.

All current approaches and proposed technologies rely on one or more pre-established organization model: EPC Global, for instance, targets exhaustibility and universality for some businesses (automotive, aeronautics, pharmaceutical, etc.).

• Severely limited code and ram space (e.g., highly desirable to fit the required code--MAC, IP and anything else needed to execute the embedded application-in, as done in ARM technology).

33

IFAC BIOROBOTICS 2018 34 Beijing, China, July 13-15, 2018

Wu Yong et al. / IFAC PapersOnLine 51-17 (2018) 31–38

2.5 Web of Things

on the Internet: a search engine that exhaustively indexes keywords found in natural language on the Web and manages to actually find (part of) the information looked for.

The Web of Things (WOT) is an inspiration from the Internet of Things where everyday devices and objects, i.e. objects that contain an embedded device or computer (i.e. MACQU), are connected by fully integrating them to the Web. Examples of smart devices and objects are wireless sensor networks, ambient devices, household appliances (i.e. speaking refrigerator), RFID tagged objects, etc.

2.6 Smart Objects Building the IOT RFID technology allows automatic identification of objects with the help of a small electronic chip (Sigrimis 1985, Piromalis, and Arvanitis, 2015 & 2016). The data stored on this smart tag can be read by wireless devices, called RFID readers that transfer to the Internet or to an Intranet, and may contain other monitoring variables forming a sentient environment that allows high resolution management (Sigrimis 1986).

Unlike in the many systems that exist for the IOT, the Web of Things is about re-using the Web standards to connect the quickly expanding eco-system of embedded devices built into everyday smart objects. Well-accepted and understood standards and blueprints (such as URI, HTTP, REST, Atom, etc.) are used to access the functionality of the smart objects. Reusing and adapting current semantic web technologies in the IOT would allow anybody to automatically publish or search object-related information guaranteed to be directly usable and significant. This association defines what will be the "Web 3.0", but we will need new approaches to succeed. We can define Web 3.0 as the conjunction of the Internet of Things with the Semantic Web (Fig. 2).

IOT develops to an enormous number of smart tags interacting with and transmitting information to each other and with decentralised and central systems. Imagine real world objects being identified by RFID and having an individual digital presence. This future dimension of technologies like RFID with a registration of every item, traced or serviced anywhere, is the IOT. Combined with GPS and mobile networks give us the capability for RTLS (Real Time Location Systems) as well as Sentient Environments. This together with Embedded Intelligence give us the potential for the upper level of automated high-resolution management. If embedded intelligence refers to agricultural crops then we will be able, using IOT, to create web services WOT (mainly on the cloud due to volume requirements) to balance our cropping actions according to SCPI concept of FAO.

Definitions World Existing Knowledge Predictions

Big Data

Ontologies

Situation

3. METHODOLOGY-AGENTS - WOT & ONTOLOGIES

Training data

Regarding the IOT, how to give objects some kind of software intelligence so that they can be able to act, react, “proact” or operate according to the context? The only way to succeed is to assess the sensor-generated information retrieved from RFID, barcodes, GPS, cameras etc. This data derived Information will be making sense in the context of the actor's end goals at a specific time, whoever the actor is: people, web services, objects, plants, etc. We then must admit that objects (i.e. greenhouses) can have goals. Therefore, giving an event a meaning is, before anything, focusing on the context of pursued objectives at a given time: i.e. a recorded low soil moisture must refer to the specific plant tolerance and salinity response, specific time or better the stage of development or even if at flowering we need to maintain a higher water soil availability, which is called URDI-Ultimate Regulated Deficit Irrigation based on the “exact context of the plant” and requires detailed knowledge of plant responses to its environment.

Theories (models etc)

New cases Predictions . 2. Data aggregation for SGH predictive modelling, Fig. management and control

Today, IOT applications only fit the requirements of closed or semi-open loops or value chain management. Each implementation is isolated and cannot easily interoperate with others. We’d rather speak of a set of "Intranet of Things", since using the term Internet is improper (we also give an example below of a closed loop for irrigation management in an intranet of the Flow-Aid consortium). However, the primary ambition of the IOT is to publish this event-related information on the Internet, in order to allow all actors involved, with object manipulation or use, to access them.

So, a Semantic web useful for the IOT is not so much trying to create links between ideas or words (semantics) but to be able to interpret the meaning of an information in a precise context: “what, where, when, how” and most importantly “why”. This requires a local autonomy in terms of perception, analysis, know-how and decision. In other words, we must put some intelligence at the lowest level (principle of subsidiarity). In the agricultural sector (Fig. 1) we seek not just the business operations but more deeply we need know

A number of techniques and standards have appeared with the aim of publishing and retrieving this information, some of which are inherited from the existing Internet static resolution systems: ONS - Object Naming system - derived from DNS by the MIT and adopted by EPC-Global. Other solutions allow for a more dynamic search, like what Google proposes 34

IFAC BIOROBOTICS 2018 Beijing, China, July 13-15, 2018

Wu Yong et al. / IFAC PapersOnLine 51-17 (2018) 31–38

the crop needs and arriving in optimal decisions with high resolution management.

be able to integrate and facilitate the generation of new services or practices: managing physical objects anywhere, anytime, anyhow; promote their reuse or sharing and facilitating the knowledge (Waldner Jean-Baptiste, 2008).

We present in the Irrigation Case a minimum of shallow knowledge (subsidiary) that is based on expert knowledge as well as a second level of intelligence that is based on encoded scientific knowledge (deep Ontologies, response or demand models etc.) as well as knowledge discovery with Embedded Intelligence. The first level expert (shallow) knowledge is at later stages, when the deep section is fully developed, not useful because the deep section empowers higher resolution but in some instances is a necessary component for dissolving conflicts or ambiguities or choosing a reasoning pathway.

Embedded Intelligence: With the advent of pervasive computer applications, the bandwidth proliferation of heterogeneous wired and wireless communication networks and the new wave on low cost environmental sensor networks have created massive collection of data while security and trust issues are yet of an obstacle. Rapid advances in sensors, web data collections and algorithms continue to fuel dramatic needs for data mining and intelligent processing. Data mining methods, enabling escience (ANNs, Bayesian networks, DBNs, decision trees, support vector machines), are becoming mature and will be used in new research for a novel way of data analysis (i.e. eirrigation). Soil sensors (Water-mark, TDR, WET, SM200, flow-aid ceramic etc) are recently advanced, and a big variety in cost and performance exists to select from, for each application. The advent of WSN is at a stage to convince that such sensor networks are already practically possible and, while Internet for everyone has become already a reality (i.e. Geomations Remote Support, through a single Satellite link and a terrestrial wLAN, or mobile internet offered by telecom companies) at a monthly cost that allows to collect data for irrigation scheduling (http://www.flow-aid.wur.nl/UK/ project). The quest in businesses, and the main objective of new research, is to gain advantage of “IOT technology everywhere”, which leads to (agri)-business intelligence. The hallmark of this new direction on knowledge discovery and exploitation is “data analysis to discover knowledge to improve decision making (for example e-irrigation, under intelligent soil sensor)”.

On the existing Web, the smart actors are human beings and their social networks: this is the underlying concept of Web 2.0 as a distinction from Web 1.0. Making objects true actors on the Web 3.0 is no more than giving them enough autonomy and intelligence, suited to the roles we want them to play. In another research effort of our team we try to create avatars of insects i.e. we use Multi-Agent system to represent the life of a whole greenhouse system (the plants, the insects, the human requirements, the physical structure etc.) and thus be able to apply integrated pest management or even integrated crop management. We need to have live digital insects to estimate risk of an infection and take pro-active actions to defeat the real counterpart. (Fig. 3).

PDD Pest Disease and Deficiencies Agent

GH

MA Mechanical Agent

MAS Sublayer specific pests

GAPs Database

GAPs Good Agricultural Practices Agent

Agent Discussion Area (Blackboard)

Although many Intelligent Data Analysis (IDA) systems have been developed (i.e. K.wiz of Thinkanalytics, Intelligent Miner of IBM, Spark of University of California, Berkeley's AMPLab, Alice of isoft etc.) most “big data” libraries remain unexploited. In science, and engineering, enormous amount of data has been generated. Molecular biology has created huge data banks and some exciting successes have been announced in bio-informatics out of data processing in protein sequencing. In soil-water-plant interactions a model may consist of many tens! of differential and non-linear algebraic equations as well as non-algebraic qualitative (decision trees) relations. Model building and model scalability are critical issues that need be addressed by the analysis of data of even a simple soil sensor.

CA Crop Agent EA Environmental Agent

Weather

UI User Interface Agent

GMH Greenhouse History Agent

User Input

35

MAS Sublayer for specific plants

Fig. 3. Overall pictorial of green-MAS Internet is ubiquitous - especially through mobile technologies - and enables us to deport intelligence on different computers. We then can associate intelligent avatars - software objects – to each physical object. Such avatars are virtual and independent software components, like “web services” and can be hosted on “cloud like” infrastructures, centralized servers, personal computers, smart phones, etc. Accordingly, the objects, through their avatars, interact and interoperate together or with other actors of the Web. By gradually increasing the level of avatar’s intelligence, we then reach ubiquitous or pervasive computing, as IOT practitioners we often draw out. In order to do this, we need design methodologies and tools for software objects conception and the patterns and naming standards (EPC, IPv6, URI, etc.). Finally, these design methodologies should

Embedded Intelligence appears in two forms: (1) Encoded existing knowledge. To be useful it must be taken from knowledge repositories (books, articles, blueprints, expertise), the best form being Models (a description of processes that has been generalized from experimental scope to a working scope and validated) and transformed to structured ontologies that have depth to allow reasoning and estimation, so decisions can be based upon. (2) Algorithms (data mining) that can analyze data records (or trans-records for multi-variable analysis) and identify models (maintenance of knowledge) or transform data to 35

IFAC BIOROBOTICS 2018 36 Beijing, China, July 13-15, 2018

Wu Yong et al. / IFAC PapersOnLine 51-17 (2018) 31–38

supporting processes and identify missing properties or parameters of a grey system. There are a considerable number of intelligent data analysis tools (ANNs, GAs, fuzzy models, SVMs, assisted by powerful methods like gradient descent back propagation, machine learning methods etc.) and create evolving systems that become smarter every day or growing up artificial life (avatars etc.), with a lately upcoming technology: the Agents style of programming that can encompass all said modern computer capabilities). It shows on Fig. 4.

agronomic research proposed irrigation-sector businesses the need to move beyond the current water use practice in Open field crops (shows in Fig. 5). We must change the highly user dependent irrigation management (with limited clock-based controllers or manual or remote mobile telephone control, product-oriented approach) and evolve towards using crop monitoring tools (IOT), remote services and information and communication technologies (service-oriented approach) with a system that can be easily retrofitted to existing installations.

Fig. 5. IOT application in Lettuce production with WOT restricted to fertigation management

Fig. 4. IOT-WOT model for green SCPI agricultural business

4.2 Concept

4. A CASE APPLICATION

In this setup we are interested in obtaining the values of the model(s) that fit the given set of continuous experimental (practice) data. Thus, the parameters (soil properties, soilwater properties, root volume, plant water requirements, plant aerial characteristics etc.) are estimated based on observing the input-output behavior of the system as shown Fig. 6.

4.1 Flow-Aid FP6 Optimal operation of commercial Fertigation systems depends mainly on the appropriate management of the available water sources, especially under saline conditions, as well as the optimal management of nutrient supply. The system developed under FLOW-AID (an EU-FP6 project), is a water management system (Anastasiou et al 2009) that can generally be used at farm level in situations where the water availability and quality is limited. This market-ready precision irrigation management system (following www.hortimed.org) is focused on new hardware and software. The hardware platform delivers a maintenance-free low-cost dielectric tensiometer and several low-end irrigation or fertigation controllers for serving various situations.

• Analyze sensor behavior (signature) and attribute to unmeasured (latent) variables. • Use point sensor measurement evolution (not just static values) to estimate state variables of surrounding (SM profile, EC, Salinity, Nutrients…. • In systems transients (irrigation event and SM evolution) have rich information about dynamic parameters. Deep learning using recursive NN and context recognition will supply much information for soil parameters and plant characteristics and instant status.

Besides improving irrigation delivery systems, an important effort must be addressed to improving water management at the field level. Unfortunately, irrigation schedules are chosen arbitrarily very often in South water scarcity European countries. Efficient irrigation schedule should be based on evapotranspiration rates, performing a simple water balance, or on root zone sensors. This proposal embraces the following FAO position: Sustainable crop production intensification (SCPI) responds to the need to increase the opportunities for crop production to address the current and future environmental threats the world is facing, and ultimately respond to the need to increase food production for the forecasted increase in human population. Hence, an important aspect of SCPI is that it looks to manage biological processes sustainably to optimize crop production.

tc Irrigation Dose

Vo

Vo (Em) Evapotranspiration

tD

VSM

tD

SM, EC measurement

t Irrigation start based on model

All the actors involved in the irrigation sector, from farmers to equipment manufacturers, share the social and political pressure to make sustainable use of resources. Advances in

Fig. 6. Relationship of irrigation and parameters 36

IFAC BIOROBOTICS 2018 Beijing, China, July 13-15, 2018

Wu Yong et al. / IFAC PapersOnLine 51-17 (2018) 31–38

Therefore, one can utilize a low frequency and a high frequency measurement of soil permittivity and try separate from that the dielectric constant which is the measure of VW content. However, there is ambiguity in this separation and the following technique has been filed for a patent.

4.3 Self-calibrated sensor Soil complex permittivity for soil moisture and soil conductivity (or salinity).

k =   − j , and   =  d +  /  0

37

(1)

As a more accurate method it can be used alone or in conjunction with a dual or multiple frequency bands or a chirp and/or phase-amplitude signature analysis method, for more unbiased and accurate results. It is understood that the regular calibration procedure is not necessary as this type of sensor provides sufficient accuracy; given a temperature sensor and temperature compensation is embedded, for both soil water content and soil conductivity (salinity).

The real part ε’ of complex permittivity k is the dielectric constant and is a measure of water content (is a measure of water molecules polarization and is higher oscillate at low frequencies), values being 2-5 for soil particles and about 80 for free water. This makes a good value at lower frequencies 20MHz because water molecules can sensitivity against soil type uncertainty.

4.4 Results on Case

The imaginary part ε” is the dielectric loss factor and is dependent on dielectric absorption ε”d and the ionic conductivity σ:

A Flow-Aid based water management system implemented IOT-WOT automatic closed loop operation was constructed as shown in Fig. 8.

(2)  =Intelligent (  −  dand )Robust 0 Sensor System: SMARTIS based on WEC soil dielectric sensor of sensors Maximumfrequency Informationand it The Intelligent: loss factor Minimum will increase with to increasing by IDA provides a means to decouple the permittivity factors, Robust: Models are Robust! Use any sensor (real) data, the down to one only, even sparse only20MHz data capture, dielectric constant at or low frequencies and and the loss most recent model to arrive to Max Information factor or the ionic conductivity at higher frequencies (Fig.7).

Web Service & Site pages

API Spec ificati ons

propriatery Format

IDA system [Intelligent Data Analysis] [Configurable for Application and for Customer]

adapt

adapt

SM sensor

EC sensor

sensor

sensor

mesh WSN

commercial controllers

Fig. 8. Flow-Aid implemented IOT-WOT automatic closed loop operation

Decoupled EC data

The data obtained is shown in Fig. 9, which were collected by PRI experiments of Flow-Aid EU project. The data recorded in comparison with conventional practice revealed the potential to saving 7% water, while improving yield by 5% and high-resolution fertigation management diminishing drainage to 0%.

Decoupling Models

SM Monitoring models

DELTA-T or Geomations

Geomations patent

excell or csv

WEC soil w_IOT sensor

Central DB

Knowledge System (SMARTIS PROJECT)

Decoupled SM data

HOST

Internet

EC monitoring models

5. CONCLUSIONS-RECOMMENDATIONS Intelligent Agriculture: The importance of intelligent data analysis arises from the fact that the modern world is a datadriven world. We are surrounded by data, numerical and otherwise, which must be analyzed and processed to convert to information, which informs, instructs, answers, or otherwise aids understanding and decision making. The quantity of such data is huge and growing, the number of sources is effectively unlimited, and the range of areas covered is vast: industrial, commercial, financial, and scientific activities are all generating such data. In agriculture this move designates our research roadpath to new dimensions of machine learning and knowledge exploitation.

Signal processing WEC sensor

under and OBI robust # 20121012244 with uptrace to EPO Fig. 7. Filed Intelligent sensor system: SMARTIS based and PCT & http://www.wipo.int/pct/ on WEC soil dielectric sensor (Filed under OBI # 20121012244 with uptrace to EPO and PCT & http://www.wipo.int/pct/)

The consumer demand for quality and safety food with environmental sustainability and the market requirements for low cost and price can only be met in a knowledge-based society. So IoT is the enabling technology to gather data-info 37

IFAC BIOROBOTICS 2018 38 Beijing, China, July 13-15, 2018

Wu Yong et al. / IFAC PapersOnLine 51-17 (2018) 31–38

from the field and AI techniques are the means to convert to knowledge and compared to Scientific Background knowledge leads to best decisions to take.

for Optimized Irrigation scheduling. Acta Hort.(ISHS) 807:253-258. http://www.actahort.org/books/807/807_33.htm Gettens J., Sigrimis N., Scott N. (1986). “Passive Activity Monitor in Livestock”. US Patent (# 4,618,861-1986). Li J, Li Li, Haihua Wang, K. P. Ferentinos, Minzan Li, Nick Sigrimis. (2017). Proactive energy management of solar greenhouses with risk assessment to enhance smart specialisation in China. Biosystems Engineering, 158, 10-22. Petropoulos G., K. Arvanitis, N. Sigrimis. (2011). Hyperion hyperspectral imagery analysis combined with machine learning classifiers for land use/cover mapping. Expert Systems with Applications, doi:10.1016/j.eswa.2011.09.083. Piromalis, D.D. and Arvanitis, K.G. 2015. Radio Frequency Identification and Wireless Sensor Networks Application Domains Integration using DASH7 Mode 2 Standard in Agriculture. International Journal of Sustainable Agricultural Management and Informatics, 1 (2), 178189. Piromalis, D.D. and Arvanitis, K.G. 2016. SensoTube: A Scalable Hardware Design Architecture for Wireless Sensors and Actuators Networks Nodes in the Agricultural Domain. Sensors (Σςιτζερλανδ), 16 (8), Paper 1227. DOI: 10.3390/s16081227. Ravinder Shergill. (2011). Service Provider IPv6 Development Experience, Senior Technology Architect. Technology Strategy, Telus CANADA. Sigrimis N., Scott N. (1985). Int’l patent (US-PO #4,510,4951985) “Remote Electronic Passive Identification-RFID”. European Patent Office #83903023.6- (154 citations). Sigrimis Nick. (2008). Automatic recording of food production and seed-to-plate traceability technology. Invited Plenary speech, SMITA CCTA International conference, Beijing China, 18-20 Oct 2008. Sigrimis N., Li Li. (2015). The PA driver today: look backwards to see forward, 6th ACPA Conference, November 16-19, Guangzhou, China. Uckelmann Dieter, Harrisson Mark, Michahelles Florian, eds. (2011). An Architectural Approach Towards the Future Internet of Things. Architecting the Internet of Things. Berlin, Germany: Springer. p. 8. DOI:10.1007/978-3642-19157-2. ISBN 978-3-642-19156-5. Retrieved 28 April 2011. Waldner Jean-Baptiste. (2008). Nanocomputers and Swarm Intelligence. London: [ISTE (publisher)]. pp. p227-p231. ISBN 1-84704-002-0.

Graphical Presentation of Treatment A VWC Fig. Fig6. 9. Graphical Presentation of Treatment A VWC (graphic (graphic in top), events (lower graph).. in top), irrigation eventsirrigation (lower graph)

At field level we will remain with electrohydraulic systems that need or supply energy and materials (machine works, fertilizers, chemicals etc). The IoT will connect the field to the knowledge world for making best decisions, which will be based on world’s first experts and knowledge centers. SOA & Ambient Intelligence will open the road to Intelligent Agriculture which we expect to: -Improve performance and cost of supply chain management -Improve resource efficiency -Enhance FOOD SAFETY, SUFFICIENCY and QUALITY -HUMAN SAFETY -Climate change adaptation and mitigation -Improve production economics (farmers’ income and NGP). 6. ACKNOWLEDGEMENTS This research was supported by the Traceability and Early warning system for supply chain of Agricultural Product: complementarities between EU and China (TEAP:PIRSESGA-2013-612659), the National Key Research and Development Program of China (2016YED0201003) and the Yunnan Academician Expert Workstation (Wang Maohua, Grant No. 2015IC16) . The authors wish to thank Acad. Prof. Wang Maohua (member of CAE), the leader of the team for solar greenhouse closed cultivation system. REFERENCES Anastasiou A., Savvas, D., G. Pasgianos, N.Sigrimis. (2008). Wireless sensors networks and decision support for irrigation scheduling, AgEng International Congress, June 2008, Hersonissos Crete. Anastasiou, A., Savvas, D., Pasgianos, G., Sigrimis, N., Stangellini, C., Kempkes, F.L.K. 2009. Decision Support 38