Agro Robotic Technology Roadmap

Agro Robotic Technology Roadmap

The future of farmbots

H.W. Sportel

13 December 2010

Title Agro Robotic Technology Roadmap Subtitle The future of farmbots

Date 13 December 2010

Place Winschoten, The Netherlands

Student H.W. Sportel

Number 1752871

University University of Groningen

Faculty Faculty of Economics and Business Programme MSc Technology Management

Course Master Thesis

1 ^st Instructor Prof. Dr. Ir. Rob van Haren 2 ^nd Instructor Dr. Ir. Ingrid ten Have, MBA

Company Productschap Akkerbouw

Associate KiemKracht

Keywords Dutch, Agro, Agriculture, Horticulture, Arable Farming, Livestock

Farming, Robot, Robotic, Technology, Roadmap, Emerging,

Trends, Drivers, Management, Innovation, Strategic, Planning,

Concepts, Stakeholder, Performance, Portfolio

Preface

Robotics become more and more important in today’s society. Robots play a role in several production processes and application domains grow in a fast pace. In the efficiency improvements of production processes, robots often are the key. The match between humans and robots is always an important aspect, also outside of the factories. We do get used to robot solutions in our society. With the children of today, growing up with so much technology around them, the future will probably yield more interesting technological robotic solutions. After the course Strategic Management of Technology of the masters programme Technology Management of the University of Groningen, I got interested even more in robotic applications. A, for me, new application domain of robots, the agro sector, was the topic of the course. Both, robotics and agriculture, are not included in the Technology Management programme.

However, the management of technological innovations like agro robotics, are an

important aspect of Technology Management. I therefore saw it as a challenge to do

research in a field, different then I was used to, hoping to contribute to the agro

robotic developments by applying technology management tools. I approached Rob

van Haren, professor at the University of Groningen and teacher of the course, in

search of a suitable thesis subject. Through him, Productschap Akkerbouw gave me the

opportunity to fulfil an internship about agro robotics. I therefore want to thank the

Productschap Akkerbouw for giving me that chance and Rob van Haren from

KiemKracht for inspiring me during this research, his enthusiasm made me curious for

research fields other than I was used to during the master. During the research, several

experts were very helpful and without there contribution, this research was useless,

thanks for that. I hope this document is as interesting to read as it was to produce and

that it contributes to solid agro robotic development and the growing of acceptation of

agro robotic solutions.

Summary

The Dutch agro sector needs radical innovation. Current machinery reached the

maximum efficiency levels. This research is focussing on the development of robotic

solutions for the agro sector now and in the future, with the goal to provide tools to

use in product portfolio decisions and to grow acceptance within the sector for agro

robotic solutions. The Dutch agro sector research is applicable in other regions. The

main research question is: How can an agro robotic product portfolio be developed,

that is a solution for emerging agro trends and improves the performance of the agro

sector? To give an answer to this question, a technology roadmap is developed to

provide agro machinery suppliers with a ‘picture of future’. The roadmap is developed

by using publications of, among others: the Universities of Wageningen and

Groningen, Dutch and European governments, the Rabobank and KiemKracht; and the

help of farmers, robotic developers and professors in agro robotics. The result is an

agro robotic technology roadmap, a document, that gives a clear image of the

emerging trends in the agro sector and their drivers; the robotic innovations and the

resources they need; linked to possible arable farming concepts. All developments are

put in a timeline of the next 15 years. Also a future vision is provided. The most

important drivers for the agro sector are: cost reduction, sustainability and the climate

change. The main agro trends, as a result, are: chain reduction and

ecological/biological farming. Important robot development domains are: Autonomy

and robustness of robotic solutions. Resources that are enablers of agro robotic

developments are: legislation to give clear direction and funding of projects by co-

operations or governments. This agro robotic technology roadmap provides insights in

future agro trends and robotic developments and grows acceptance for agro robotic

solutions. This will contribute in the decision making in future product portfolio

choices in agro machinery suppliers businesses.

Content

Preface... 1

Summary... 2

Content ... 3

1 Introduction... 4

2 Research Definitions ... 6

2.1 Context ... 6

2.2 Problem ... 7

2.3 Goal... 7

2.4 Model... 8

2.5 Questions... 10

3 Research Methodology ... 12

3.1 Technological Innovation Management... 12

3.2 Technology Roadmapping Framework... 16

3.3 Research Methodology... 23

4 Stakeholder Analysis... 24

4.1 Research Methodology... 24

4.2 Stakeholder Analysis... 27

5 Technology Roadmap... 31

5.1 Drivers... 31

5.2 Sector... 36

5.3 Technology ... 45

5.4 Resources... 59

5.5 Concepts ... 61

6 Research Evaluation... 68

6.1 Theoretical Framework ... 68

6.2 Tool Applicability ... 69

7 Conclusions and Recommendations ... 72

7.1 Research Questions ... 72

7.2 Research Conclusions ... 75

7.3 Recommendations and Further Research... 77

Acknowledgements ... 79

Reference List... 81

Further Reading... 85

1 Introduction

Farms and forests cover most of Europe’s land and are vital for our health and economy (European Union, 2010 ¹ ). Agricultural systems are at the heart of developing countries’ economies and family livelihoods and the development is explored within a dynamic context (Kanyama-Phiri et al, 2008). Different aspects as limited resources, pandemics, climate change and global population growth put pressure on innovative agro solutions (Ash et al., 2007). The agro sector is mainly conservative and current suppliers of agro machinery are therefore still focussing on improving current techniques; innovative solutions (disruptive technologies or radical innovations) are marginally at place (Blay-Palmer, 2002). Although Dutch agriculture is on a high innovative level with biological farming, smart farming and projects for achieving high farming efficiencies, continuous innovation is needed (Verkaik, 1998). One of the technology areas that can give the innovation process a boost, is robotics. Robotic innovation is developing rapidly and the scope of applications is growing. Robot technologies are already used in milking machines ² and some weed robots ³ . But the development of robotics is diverse and future technological developments are difficult to predict. Linkage to the agro sector (developing agro robotics) is therefore complex.

Also other factors play a role in agro robotics. Legislation, politics and funding, but also the acceptance of robot technology is important in developing agro robotic solutions.

Suppliers of agro machinery play a great role in innovation projects and the development of agro robotic solutions. The fundamental thesis of this research lays in the question how suppliers of agro machinery can improve their business performance by pursuing a agro robotic strategy. Clear answers can enable innovation projects by investments of agro machinery suppliers. The development of a agro robotic product portfolio that fits the vision and the needs of the agro sector is important in improving

1

http://europa.eu/pol/agr/index_en.htm

2

http://library.wur.nl/dps/diensten/dossiers/melkrobots.html

3

http://www.wetenschap24.nl/video/bekijk/ruud-de-wiedrobot.htm

the business performance of agro machinery suppliers. Knowledge about product portfolio requirements by analysing the market and the technological possibilities is therefore of great value for agro machinery suppliers. This research is therefore focussing on how future developments of agro robotics can be guided, so that it can be used by agro suppliers for agro robotic product portfolio development purposes.

Important in this research (besides a literature study) is the vision of the agro sector and the development of agro robotic concepts by using future robotic developments.

This is the foundation of the assignment of resources. One of the pursued effects of this research is to create consensus about the future agro robotic solutions and contribute to agro robotic acceptance.

The report is structured as followed. After this brief introduction, the research

definitions are discussed in the second chapter. This will contain the context of this

research with the chosen problem and the goal. Also a research model is presented in

this chapter, along with research questions. The third chapter is presenting a research

methodology for the management of innovations and in particular within the agro

robotic concept. A framework is presented in this chapter, resulting in a format which

is used as roadmap for agro robotics, combined with a research methodology which is

used in the this research. Chapter four is discussing stakeholders by presenting a

methodology for stakeholder analysis and using that to make a first analysis of the

stakeholders involved in agro robotics. The agro robotic technology roadmap is

developed in chapter five by using the categories drivers, sector, technology, resources

and concepts. This research consists also out of a evaluation of the tools that are used,

in chapter six, followed by the conclusions of the research and the answers of the

research questions in chapter seven.

2 Research Definitions

Robotic development evolves rapidly and is diverse; great investments in technology are needed. To improve the probability of success and minimize the (financial) risks of developing robotic solutions, product engineers for robotic solutions in the agro sector need clear statements of future agro trends and future technological possibilities in developing concepts. Clear pictures of future developments are difficult to achieve when addressing technological innovations, but by using multi angle perspective on future developments, still solid statements of the future agro robotics can be developed. The research definitions in this chapter are therefore important to start with. To structure this research, a research model is made. First the context is described in paragraph one. That is followed by the problem statement in the second paragraph. The third paragraph is describing the goal of this research. All these aspects are graphically structured in the conceptual model in paragraph four, which leads to the research questions in the fifth paragraph.

2.1 Context

The global agro sector is subject to great differences, intercontinental but also between countries and even regions. Soil, temperature and water are the main production factors in agriculture (Dan Richter (2008) Duke University North Carolina) ⁴ . These factors differ between regions in the world. Because of these regional differences and the difficulties in researching large areas, this research is only focussing on the Dutch agro sector as a representative for the industrialized type of agriculture, especially arable farming. The agro sector is defined as the science, art and business of cultivating soil, producing crops and raising livestock; in other words:

farming. ⁵ This includes: (greenhouse) horticulture, arable farming and livestock farming. Within the agro sector context, the focus or this research is on (concepts of)

4

http://www.peakoil.nl/2008/02/04/invloed-klimaatverandering-op-voedselproductie/

5

http://www.thefreedictionary.com/agriculture

robotic technology, defined as the science or study of the technology associated with the design, fabrication, theory and application of robots ⁶ . Specifically, how those concepts of robotics can contribute to the enhancement of performance of the whole agro sector. The link with agro machinery suppliers is made by providing clear future directions of the technologies and market trends.

2.2 Problem

Robotic innovations are diverse and develop rapidly. Development requires many different specializations and is expensive. For agro machinery manufacturers, it is not completely clear how robotics can contribute to performance enhancement of the agricultural sector, now and in the future. Different technological approaches exist, making the field of robotics unclear. The acceptance of robotic solutions in the agro sector is still not optimal, making robotic solutions complex to develop. These circumstances make it difficult for agro machinery manufacturers to develop a profitable business strategy with an agro robotics product portfolio.

2.3 Goal

This research has several goals, for the agro sector as well as for the suppliers of agro machinery. The first goal is to develop concept requirements, based on the agro sector vision on robotics and the robot technology vision. These requirements can be used to develop concepts for a robotic product portfolio. For the agro machinery suppliers it is also important to get a clear view of the resources needed to achieve these developments. Another goal is the growth of robotic awareness and acceptance of robotic solutions in the agro sector. The main goal is to develop a framework that can be used to guide the future of agro robotic developments. A framework that can be used to manage these technological innovations.

6

http://www.thefreedictionary.com/robotics

2.4 Model

The research, as discussed above, can be drawn in a conceptual model of the problem situation. This model is presented in figure 1. Of course, this research is focussing on the Dutch agro sector and in specific the use of robotic applications, but is applicable to the agricultural developments of the western society. The models that can be drawn of the situation, are discussed below. The research questions that are based upon the models, are discussed in the next paragraph.

Figure 1: Conceptual Model - Problem

The drive of this research is the business strategy of suppliers of agro machinery, that enables agro robotic solutions. The goal of the machinery suppliers is an improvement of their business performance, by adapting an agro robotic business strategy. A result of an agro robotic business strategy can be found in their product portfolio. The product portfolio needs to fit with the agro sector.

The conceptual model of the problem (figure x) starts with the management question whether the agro supplier business wants to adapt an agro robotic strategy. If so, the business needs analysis of robot technologies and an analysis of the agro sector (hence: both aspects of agro robotics). The results can be summarized in agro robotic requirements, needed for the development of agro robotic concepts. These concepts are used in developing an agro robotic product portfolio. The goal is, of course, that the agro robotic product portfolio fits the agro sector. If not, new market analysis is needed to research the vision of the agro sector. If the product portfolio fits the vision of the agro sector, the search for new robotic technologies continues, because the evolution of technology will continue in a fast pace. Of course, even when the product portfolio fits the vision of the agro sector, keeping the analysis up to date is important.

When the product portfolio fits the agro sector, the new agro robotic solutions are used to improve the performance (in a wide perspective) of the agro sector. The assumption is that this situation is enabling the improvement of the business performance of the agro machinery suppliers (e.g. by sales of products).In this research, not the whole process as described above, is analyzed. In the fuzzy frond end of innovation, where this research is done, only the first phase, containing the analysis of robotic technology and of the agro sector, is important. It is important in the development of agro robotic concepts/prototypes and eventually the development of an agro robotic product portfolio. How this area is researched is shown in the conceptual model of the research in figure 2 on the next page.

As already discussed in the first problem model, the research consists out of two

aspects: robotics and the agro sector. The agro sector can be divided into two areas:

emerging trends and sector performance. New agro robotic solutions can address both areas to be successful. The performance of the sector is about the current situation with the current product portfolio. The emerging trends are about the future trends in the agro sector. An important issue in emerging trends, are the drivers which enable the trends. The other aspect (robotics) can be divided into the current robotic technologies and other emerging technologies, which can be used in an agro robotic context. In this aspect, the resources which are needed to develop the technologies are an important issue. In the innovation process, the management of stakeholders is a difficult but important issue. Therefore, a stakeholder analysis is also part of the development of agro robotic requirements.

Robotic Technology Analysis

Agro Sector Analysis

Agro Robotic Requirements

Sector Performance Emerging Trends

Drivers

Robotic Development Emerging Technologies

Recourses

Agro Robotic Innovation Stakeholders

Figure 2: Conceptual Model - Research

2.5 Questions

The above discussed research models (based upon the context, problem and goal) can

be translated into research questions. The first question is the main research question,

which is the fundamental question of this research, and it is stated as followed:

1. How can an agro robotic product portfolio be developed, that is a solution for emerging agro trends and improves the performance of the agro sector?

By answering this research question, extra knowledge can be added which suppliers of agro machinery can use for developing an agro robotic product portfolio, which fits the agro sector and therefore helps to create acceptance of robotic solutions in the agro sector. This will contribute to the overall goal of improving the business performance of agro machinery suppliers by adapting an agro robotic strategy. But to answer this main research question, the following research sub questions need to be discussed:

a. What is the performance of the agro sector, and how can it be improved?

b. What are emerging trends and what are their drivers?

c. What robotic developments can be useful in the agro sector?

d. What are emerging technologies and what resources do they need?

e. What stakeholders are important and how can they be addressed?

f. Can a framework be developed to manage this technological innovation?

The first and the second research sub question is aiming on obtaining knowledge about the agro sector, while the third and fourth research sub question is aiming at expanding knowledge of emerging robotic technology developments. Both aspects combined, result in the main part of answering the main research question. The fifth research sub question is about analyzing stakeholders. This is an important issue since stakeholder(group)s could have influence in future developments. Maybe the most important research sub question is the sixth, where all results are put together in a search for a framework that is useful in guiding agro robotics.

A discussion on how the above research questions can be addressed, will be made in

the next chapter, research methodology. In that chapter, the different possibilities will

be elaborated by developing a theoretical framework based upon a literature survey.

3 Research Methodology

In this research it is important to make clear what the future agro robotic developments can be, so that the agro machinery businesses can develop a product portfolio that fits the vision of the agro sector. This implies that two aspects are important, the agro sector and the robot development. Therefore, a research methodology is needed that can address these future technological innovations and provide tools for concept developers. In this chapter a suitable methodology and the conditions to make the methodology work, will be discussed. In the first paragraph, the management of technological innovations is addressed, within the context of this research. The framework of the Technology Roadmap is presented in the second paragraph.

3.1 Technological Innovation Management

In new product development, several tools can be used to structure the development process and optimize the results. Ulrich et al. (2008) describe different phases in the development process and give different tools to use in each phase; tools like project planning and lean development. Also TRIZ ⁷ , Analytic Hierarchy Process (AHP) and Technology Roadmapping (TRM) can be used in multiple New Product Development (NPD) scenarios (Lee et al., 2008). But most of these tools are focussing at generating concepts or improving development routes with a short time horizon, focussing on incremental innovations. However, the research for agro robotic solutions is dealing with long term solutions, high uncertainty an unknown markets. Circumstances that are typical for the so called Fuzzy Frond End (FFE) (Crawford et al., 2006). In this stage (figure 3 on the next page), problems arise with issues like how can future technological developments be forecasted? Is there a framework that can help to plan

7

http://www.nieuweproductenbedenken.nl/default.asp?pid=840

and coordinate those developments in technology? How can consensus about agro robotics future developments be reached?

Figure 3: Fuzzy Frond End

Tidd et al. (2005) mention three main phases of an innovation core process: Search (1), where new technologies are used in developing concepts for new markets; Select (2), where the choice of concepts to continue with is made and those concepts are developed further; and Implement (3), where the concepts are in the early production phase. For businesses, the key in all of these phases is learning, so that the way that innovation and development processes are managed, can be improved. The decision of managers whether to invest in early-stage technologies, is based on qualitative measures (Phaal et al., 2009). In the early uncertain stage of technology development, they lack accuracy. It is therefore important to focus on quantitative value. In an innovative context, Koen et al. (2002) describe the FFE phase by using the New Concept Development (NCD) model (figure 4 on the next page), with three phases:

Opportunity (1), where a business or technology gap is recognized and explored; Idea

(2), where embryonic new products or services are developed; and Concept (3), where

well defined written and visual concepts are developed. The model is adapted from

Koen et al. (2002) with minor changes (the stage idea generation & idea enrichment is

split in two separate stages, both in the idea phase. The idea selection stage moved to

the concept phase) to make the stages fit better to the phases. This research is

operating within the opportunity phase, the blue area in the model. Within this phase,

several tools can be used to analyze opportunities, generate ideas and develop

concepts. These tools are discussed shortly below.

Figure 4: Concept Development Model

Koen et al. (2002) mention specific tools to use in the opportunity phase. These are tools concerning trend analysis, market research, scenario planning and roadmapping.

These tools are also discussed by Tran et al. (2008) for technological forecasting

purposes. They also discuss other tools like risk assessment and decision analysis. As

discussed by the authors in their article, each tool has a different goal and can be used

to fulfil different purposes. The tools trend analysis, market research and roadmapping

are the most suitable for fulfilling the goal of this research; defining the future

direction of agro robotics. In a further stage (i.e. stages that follow up this research)

other tools like scenario planning, risk analysis and decision analysis can be useful. A

trend analysis can be divided in a technology trend analysis and a market trend

analysis. The second can be completed with a total market analysis. Depending on the

format used, by roadmapping, both technology and market can be put together and

linked by concept development. This gives a new perspective, since the it does not consist out of two single analysis, but the link between technology and market. The goal of trend analysis is to achieve clear statements about future trends and/or future technologies. The goal of roadmapping is the development of concepts, using future trends and technologies.

Drew (2006) argues that the uncertainty of disruptive technologies make the strategic planning to focus more on learning than on implementation. To create a shared vision for future designs and required component technologies, Technology Roadmapping (TRM) is an essential business tool (Fischer, 2006). Lee et al. (2008) also used TRM for project portfolio management and argue that TRM can take the long-term strategy of a firm into account, where other portfolio management methods failed.

In the idea phase, tools like new TRIZ (or SIT) are helpful in processes of idea generation and enrichment. The starting point of this is always the current product.

Improvement can be done on three levels (i.e. past, current and future) but the techniques are focussed on creativity; future trends in the market or emerging technologies get less attention than needed for this research (Altshuller, 1979).

Furthermore, Souchkov (2007) argues that TRIZ can be used perfectly for developing breakthrough solutions. In that way, TRIZ can be used perfectly by manufacturers of agro machinery when developing agro robotic concepts to improve their product portfolio. Nevertheless, in this research, where the search for market trends and emerging technologies is the central issue, TRIZ is not the optimal tool.

The main research question is focussing on the development of agro robotic solutions

for emerging trends and sector improvement. For answering the main research

question, a technique like technology roadmapping is more useful to get insights in

emerging trends and technologies (Koen et al., 2002). The next paragraph will

elaborate more on technology roadmapping techniques and formats.

3.2 Technology Roadmapping Framework

The first step in developing a TRM is choosing a suitable format. A technology roadmap can be developed out of different process design choices. One of the first steps in developing a suitable technology roadmap for agro robotics, is the choice between a roadmap on company level or on industry level. On company level, the technology aspects, development and needed resources, but also market aspects are researched on company level. This is an internal process, that is part of the planning activities of the company. A roadmap on industry level consists more than one company and collaboration is needed to get solid information about the future of the sector (Garcia et al., 2007). An example is the Semiconductor Industry Assosiation (SIA) Semiconductor Technology Roadmap ⁸ (examples of roadmaps of Samsung ⁹ in figure 5 and 6 on the next page). To make a solid image of the future agro sector, an industry level roadmap is chosen.

Lichtentaler (2008) argues that, while developing a roadmap, companies can benefit from open innovation because of the development of industry standards and create the freedom to operate. Open innovation therefore has a positive effect on the probability of success of innovation projects. This is especially the case in situations where companies find it difficult to identify possible applications in other contexts for their potential technological solution, because the answer can often be found in industries that are completely different from their own product business. Robotic development can be seen as a technology push process, developing robotic technologies first and define application purposes later (especially in agro robotics).

Open innovation at industry level could therefore help in developing sound agro concepts. The Autonomous Sugar Beet Harvest project however, showed that an open innovation structure is not yet fully accepted by agro machinery suppliers. Such a structure will therefore not been used in this research.

8

http://www.itrs.net/

9

http://www.samsung.com/global/business/semiconductor/products/fusionmemory/Products_MCP_pkginfo.html

Figure 5: Technology Roadmap (1

example Samsung)

Figure 6: Technology Roadmap (2

example Samsung)

Another choice that Garcia et al. (2007) discusses is one between a product technology

roadmap and an emerging technology roadmap. Where a product technology

roadmap is focussing on products (and their technologies) as well as markets (and their

drivers), an emerging technology roadmap is only focussing on one emerging

technology (in a product) and the future developments and applications of that

technology. Emerging technologies are an important aspect in agro robotic

developments, however, in this research the focus is on combining the market aspects

with several robotic technologies so that requirements for concept development

(products) can be made. Therefore, the product technology roadmap is used.

Figure 7: Roadmap Formats

Now the choice of using a product technology industry roadmap based upon open innovation developments, the format of the roadmap must be determined. Lee et al.

(2005) discuss eight different formats (as shown in figure 7, a-h). Without further

discussing these formats, in this research the central issue is future agro robotics. This

dynamic external research for technology trends implies that a technology trend

roadmap (figure 7, image h) is the most useful. In that format, the factor time can be

deployed as well as the different aspects (market, drivers, technology, resources and concepts), which is perfectly applicable in this agro robotic research.

For this research, the agro sector with their trends, drivers, performance and problems is very important. This will be researched in the subject market. Within the market, a distinction can be made between the different business types (i.e. horticulture, arable farming and livestock farming). One can imagine that development of trends and performance aspects differ from each other. Therefore, the market analysis is taken the aspects as well as the different business types into account. An issue in defining the market aspects, is the timeline used in the roadmap. For implementing new agro robotic concepts, new strategies have to be deployed. Rotation of cropping can be up to seven years in some circumstances (Snapp et al., 2008), like biological farming rotations. The average crop rotation in farming businesses however, is between four and six years, which means that the current short time planning horizon of farming businesses (especially in the arable farming) is also at least between four and six years.

This has effects on planning horizons and investment projects. Because of the differences within the agro sector, the subsectors horticulture, arable farming and livestock farming are used in the roadmap.

The other important subject in agro robotics is, of course, robotic developments. The robotic developments are diverse and several technologies can be used in robotic solutions. Neuro technology, bio technology, but also software for image recognition and autonomy can be important aspects in robotic developments. To create an agro robotic vision, the robotic vision is based upon robotic developments and emerging technology trends. Sato et al., 2009 developed a technology roadmap about the role of robotics in society (see figure 8). This roadmap is focussing on robotic developments over the next 50 years. Two interesting aspects in this roadmap are missing however.

First, a timeline is missing, the development of robotics in time is very interesting, even

indispensable, when concepts for agro solutions are to be designed. Second, a

timeframe of 50 years will increase the level of uncertainty whether these

developments will ever occur, and at what point in time. Robotic developments are fast and diffuse. Long time frames cause high uncertainty in roadmaps. Agro planning horizons are short time on average five years, long time and a sector vision are far in the future. To combine both the agro planning horizon and the robotic technology developments, the following timeframe is chosen: present (<2011), short term (2011 – 2015), mid term (2016 – 2020), long term (2021 – 2025) and vision (>2025).

Figure 8: Robotic Roadmap (Sato et al. 2009)

In robotics, there are very different approaches of the technologies ¹⁰ that are used.

The categorisation is therefore somehow subjective. With information from the DFKI robotic research centre in Bremen and lecturers robotics from the University of Groningen, the following approach is used in defining important technologies for agro

10

http://mediatheek.thinkquest.nl/~ll106/Knowledge/Technologies/main_D.html

robotic concepts and the development of a roadmap. The first important aspects are software and hardware, quite basic technologies of robotics in general. The use of sensors and the communication are also important aspects in developing robotic solutions. Also the energy aspects play an important role in developing (outdoor) robotic solutions. Within the agricultural context, the use of geo positioning is very important and therefore added to the technology issues within the context of agro robotics.

The linkage between technologies and trends is made by developing concepts.

Concepts can be divided in four domains, based upon ten brainstorm sessions ¹¹ , organised by Kiemkracht, together with robotic developers and farmers across the Netherlands. The result is the use of four different concept domains, knowingly inspection & control, crop inspection & management, harvesting and post harvesting.

The arable farming concepts in this roadmap will be categorised within one of these concept domains.

The above discussion leads to the choice of an industry level product technology trend roadmap, consisting out of five domains: the trends (1) of the agro sector and their drivers (2), robotic technologies (3) and the resources needed (4), linked to each other by arable farming concepts (5). The agro sector is divided is four domains (i.e. total agricultural sector (1), horticulture (2), livestock farming (3) and arable farming (4)).

The robot technology is divided in the six main components (i.e. software (1), hardware (2), sensors (3), communication (4), energy (5) and geo positioning (6)). The arable farming concepts are divided in four domains (i.e. inspection & control (1), crop inspection & management (2), harvesting (3) and post harvesting (4)). The timeline will used is: present (<2011), short term (2011 – 2015), mid term (2016 – 2020), long term (2021 – 2025) and vision (>2025). Research using this format (figure 9 on the next page) will contribute to the search for an optimal agro robotic product portfolio that fits the vision and performance of the agro sector, using robotic technologies.

11

De route naar een één-rijige bietenrooier (SmartBot Brainstorm, 2009)

Figure 9: Agro Robotics Technology Roadmap Format

3.3 Research Methodology

For the completion of the agro robotic technology roadmap, several research methodologies will be necessary. Using different publications (Stichting Natuur en Milieu, Rabobank, University of Wageningen and Massey University) concerning emerging agro trends and the performance of the agro sector, a first draft of the agro sector will be deployed. In this phase, also early brainstorm sessions with framers, done by Kiemkracht, will be used. The result will be formed in a short list of emerging agro trends. A discussion with several farmers in different agro businesses about this short list, will result in a solid image of the emerging trends and their drivers. This will contribute in answering the first two (i.e. a and b on page 11) research sub questions.

A comparable approach is used for answering the second two (i.e. c and d on page 11) research sub questions. Using publications about emerging (robot) technologies, the foundation is made for the technology aspect of the roadmap. Further developments are discussed with DFKI Bremen and other experts, resulting in a clear image of the emerging (robot) technologies and their possible resources, consistently using the format of page 22 to update and improve continuously.

As discussed before, an important aspect that can have major influence in the future developments within the agro sector, are stakeholders. The different stakeholder dynamics can cause roadmaps to develop in different directions. A clear view on how these stakeholder dynamics can be managed is needed to make the roadmap a solid tool for portfolio development. For answering the fifth (i.e. e on page 11) research sub question, extensive literature research is done and tools like stakeholder mapping and stakeholder analysis are used.

All aspects above are researched and the results are used to develop some kind of

framework that can help in developing an agro robotic roadmap. This will answer the

last research sub question (i.e. f on page 11) and provide tools for further research and

development of agro robotics.

4 Stakeholder Analysis

Besides the technological aspects of developing a technological roadmap, other social factors are to keep in mind. An important aspect in the development of agro robotic solutions and the use of products within the portfolio, is acceptance by users. Other valuable parties are governments and investors, because of their great influences on funding, regulation and direction of developments. Furthermore, other agro business and governments are to be effected by new agro robotic solutions, creating new possibilities for new world markets and change current power relations. This means that stakeholder(group)s are important in the development of an agro robotic technology roadmap, since they have great influence. The dynamics of the stakeholders can cause major issues and stakeholder analysis should therefore be a substantial part of the roadmap development.

This chapter consists out of two parts. The first part is to define a research methodology for the stakeholder analysis, this is discussed in paragraph one. The second paragraph will elaborate more this methodology and starts with performing a fist stakeholder analysis for agro robotics. This will be discussed in paragraph two.

4.1 Research Methodology

The last decades, since Freeman published his book about a stakeholder approach in

strategic management (1984), the importance of stakeholders is more acknowledged

and thorough analysis-tools are developed (Mallin, 2007). The defining of stakeholders

can be done by using the terms internal (a), when the stakeholders are part of the

system, external (b), when the stakeholders are outside the system, but have a clear

link, and interface (c), when stakeholders can influence the system by laws and

regulation. Another definition can be made in the distinction between primary (1),

when the stakeholders are directly influenced by the system, and secondary (2), when

stakeholders are indirectly influenced by the system (Donaldson et al., 1995). Within those categories, key stakeholders can be pointed out, which can have large influence upon or importance within the system.

A familiar classification of stakeholders is based on power of stakeholders (a), legitimacy of stakeholders (b) and the urgency of stakeholders (c). The result is a stakeholder salience level, implicating the level of management attention the stakeholder needs (Mitchell et al., 1997). Fletcher et al. (2003) defined a process for mapping stakeholder expectations based on value hierarchies and Key Performance Areas (KPA). Savage et al. (1991) offer a way to classify stakeholders according to potential for threat and potential for cooperation. Another process of identification is by an assessment of awareness, support, influence leading to strategies for communication and assessing stakeholder satisfaction, and who is aware or ignorant and whether their attitude is supportive or opposing (Turner et al, 2002).

S up po rt

In te re st 1 0

10

Figure 10: Stakeholder Classification Example (power 2/10; interest 8/10; support 8/10)

When the Mitchell theory is simplified, a common used classification can be made by

classifying stakeholders in three categories; power, interest and support (see figure 10

on the previous page, where an example is made where power is classified with a 2/10 score, interest with a 8/10 score and support also with a 8/10 score). By classifying each stakeholder(group) on these three aspects, a clear image on how to manage these stakeholders can be made. These three aspects are very important in agro robotic (or any technological innovation project). Knowledge about supporting stakeholders is important, they can be used to boost the project and influence high power stakeholders. The contribution of stakeholders with high interest during the project is the key for developing successful concepts. The slightly modified Mitchell model is therefore perfectly useful in this agro robotic research context.

Vos et al. (2006) argue that for improving the outcomes of innovation processes, the project management have to continuously decide which stakeholders are important and how these stakeholders can be addressed. An important aspect in the stakeholder developments during innovation is stakeholder dynamics (Postema, 2010). This means that during the process, stakeholders could change. New stakeholders could appear and current stakeholders could disappear. But also the classification of stakeholders could change, the power difference of change in support could easily change and the forming of (other) stakeholder groups could be a fact. This implies that the approach of stakeholders and stakeholder management is also subject to change. Therefore, an important factor in stakeholder analysis is that it is a continuous process, and an iterative cycle.

In this agro robotic context, an iterative stakeholder analyses process is recommended.

This is because of the process of development agro robotic solutions is a path of

radical innovation, which enables stakeholder dynamics. The iterative cycle of the

stakeholder analysis (see figure 11 on the next page) starts with indentifying the

stakeholders (1), then the stakeholders are classified (2) by using the three

dimensional model (i.e. power, interest, support) so that adequate stakeholder

management (3) actions can be considered. To keep managing the stakeholders in the

most sufficient way, continuous stakeholder identification and classification is needed

to refresh the approach of stakeholders and conduct effective stakeholder management. A first start with the stakeholder analysis is made in the next paragraph.

Figure 11: Stakeholder Analysis Process

4.2 Stakeholder Analysis

In this paragraph a stakeholder analysis is performed. Of course, only a first cycle of stakeholder analysis can be done at this stage and stakeholders can only be analysed in general. The stakeholder analysis starts with identifying the stakeholders for the agro sector, then the identified stakeholders are classified. These results are the bases when defining a stakeholder management plan. This plan will also contain future stakeholder analysis recommendations.

Stakeholder identification

For the stakeholder analysis, an extensive stakeholder list of agri-infotech ¹² is used as starting point. The stakeholders of the agro sector (see figure 12 on the next page) is divided in thirteen stakeholder categories from consumers to government agencies.

Within these categories, several stakeholder groups are identified. An elaboration of the stakeholder identification can be made when the robotic innovation stakeholders are also taken into account. However, the most important stakeholder issues in robot

12

http://www.agriinfotech.com/

innovation are based upon the sector is its used in. In their research for robot development in health care, Boxsel et al. (2008), the stakeholder identification is solely based upon the sector it performs in (i.e. health care). In this research however, the focus is on robot innovation in Dutch agro sector. This could mean that the people who are involved in the insufficient Dutch innovation climate (Roel Smit, 2007 ¹³ ) are also important stakeholders. Especially because of the stakeholder dynamics, the identification of stakeholder(groups) will change over time and more extensive stakeholder research at this point in time is not included in this research and is important when the project evolves in time and different concepts/prototypes are developed.

Sellers Buyers

Ind ependent individu al operations Indepen dent individual buyers

Small aggregators Small-independent retailers

Whole sellers Institutional bulk buyers

Expo rters and Importers Whole sellers

Corporate sellers Exporters and Imp orters

Retail Chains

Consumers Knowledge Management Service Providers

Ingredients and raw produce consumers Indepen dent individual consultant - advisors Processed bulk produce consumers Not for profit - volun tary organizations Retail packed produce consumers Contract farming organizations

Group promoters - d eveloper - managers Aggregators

Network Marketing Processing Facilities

Ind ependent individu als Micro enterprises

Unorganized groups Small an d medium enterprises

Organized groups Large processors

Regulatory - Quality Certification agencies Non-Government-Organizations

Accreditation agencies Laybour groups

Quality regulatory agencies Environmental groups Quality certification agencies

Producers Logistics and Linkage Service Providers

Ind ependent individu al farmers/producers Packaging and labelling

Groups o f farmers Transp ort

Co-operatives Warehousing

Contract farming agency Handling

Corporate farming agency Procurement management

Cold chain management

Network Suppliers Banking, credit and investment services Suppliers of inputs Insurance and risk man agement services Suppliers of tools and equipments IT and communication services Suppliers of technology and services

Governments Others

Product boards Trade promotion body-agencies

Water boards Research organizations

National governments Academic institution s

European governments Techn ology development agencies Extensio n agencies

Quality management organizations

Figure 12: Stakeholder Groups (based upon Agri-Infotech stakeholder analysis)

13

http://www.vno-ncw.nl/publicaties/

Stakeholder classification

For each stakeholder(group) from the identification list, a classification can be made.

These classifications are based upon the presented terms power, interest and support.

Stakeholder classification can only be done for specific agro robotic innovation projects. Each project (e.g. concept/prototype) is linking different technological features to different market aspects, purposing different solutions. Stakeholder classification will therefore be different in any case, depending on factors like market segment, technological features, aimed solution and disruptiveness. The classification of stakeholder is therefore very important. However, at this point in time, a thorough stakeholder classification is not adding value to the research. Hence, the stakeholder dynamics have great influence in the way stakeholders are identified and classified.

However, at this point in time, statements can be made about important stakeholders developments that enable or delay agro robotic developments. Great enablers of agro robotic developments are research groups of the University of Wageningen (by developing agro robotic solutions) and agricultural product boards (that support agro robotic research). Other important stakeholder groups are large manufacturing companies of agro machinery (i.e. Grimme) that have high interest in the developments of radical agro innovation. And last but not least, the farmers. They need to adapt agro robotic solutions, to make the developments successful.

Stakeholder management

Strategies of stakeholder management are important because they enable maximum project results. Especially in case of innovation projects, stakeholder management plays an important role in obtaining resources, development of technologies and acceptance of concepts/prototypes. The management of stakeholder can be done on three areas (i.e. power, interest, support). The bases of management of power stakeholders lays in strategies that also enables (a) high support or (b) high interest.

Stakeholders that can only be classified as high power (neutral on support and neutral

on interest) can be dangerous for the project, since they are very dynamic and the

neutral support levels can easily change to low support. One of the reasons for that phenomenon is that stakeholders with low support search for powerful stakeholders as allies to strengthen their point of view. The bases of management of high support stakeholders, lays in enabling strategies. This means that a project needs supporting stakeholders, to strengthen their own position. Especially in innovative projects, support is very important in creating acceptance and continuation. Enabling strategies can also focus on obtaining higher power levels. Stakeholders on a high interest level, often search for explicit support levels (either support or no support). The bases of management of interest strategies lays in growing commitment, aiming for high support levels. The key in stakeholder management is to create space for stakeholder dynamics, keep communication open and have a pro-active attitude. For each innovation project within agro robotics, stakeholder management will be different.

Within the context of technology roadmapping, stakeholder management is very important. The roadmap is a development path of technologies and trends in the future. The development path consists out of different steps or phases that are necessary to complete the picture of the future. Every robotic technological innovation or changing trends in the agro sector means shifting stakeholder interests, support and power. Even new stakeholder(groups) can arise and others could disappear. It is not only the shift of stakeholders when the process is moving to different phases, it is also involvement of stakeholders that may enable the entry of different phases in the roadmapping process. Stakeholder management actions can contribute the speed of developments within roadmapping contexts. The stakeholder dynamics that occur during the evolution of the roadmap, need consistent attention and it is therefore important to address stakeholders during the technology roadmapping process.

However, it is only applicable at concept development level (since the complexity and

broad view of the total roadmap). To connect possible (future) stakeholder(group)s to

agro robotic development, open innovation structures might be a solution, where high

support and/or high interest stakeholders (i.e. parners) grow a consensus about agro

robotics.

5 Technology Roadmap

In this chapter, the agro robotic technology roadmap is developed. The roadmap format as purposed in figure 9 on page 22 is used. As can be seen in the figure, the format consist out of five timeline objectives (i.e. present, short term, mid term, long term and vision) and five aspect objectives (i.e. drivers, agro sector, arable farming concepts, robotic technology and resources). These five aspects are discussed below.

The structure in the discussion of the different aspects, is lead by elaborating on the different trends within that aspect. In the first paragraph, the drivers are discussed, followed by the trends in the agro sector in paragraph two. The robot technology are discussed in paragraph three and the fourth paragraph is addressing the resources needed. Current concept development and implications for the development of future concepts in agro robotic business context, are discussed in the fifth and last paragraph.

How the different drivers, trends, concepts and resources are situated in the roadmap format, can be seen in figure 27 on page 68, where the result of the agro robotic roadmap is shown.

5.1 Drivers

Within the context of developing robotic solutions for the Dutch agro sector now and in the future, different drivers can be identified. These drivers are important since they are key enablers for trends in the agro sector. Identifying these drivers makes project management possible that structures innovation developments to meet (future) trends in the sector. For each context, different drivers can be identified. For agro robotics, the main drivers are discussed below.

Cost Reduction

The main driver for current robotic agro innovations are cost reductions (Yves de

Groote, 2009). Cost reduction due to higher efficiencies of agro processes or the

substitution of expensive human capital. Achieving cost reduction within one of these factors is the most important goal of the current prototyping activities of agro robotics.

With the declining human agro capital, new solutions have to be found. The same need for new agro solutions with robotic technology is developed to compete with foreign agro businesses. The goal of agro robotic prototypes is to increase flexibility by reducing the need for the human capital constraint. This makes cost reduction an important driver on the short term.

Fair Competition

For decades, fair competition is an issue that has great influence in government policies concerning agriculture. Different circumstances in different countries and continents are critical in difficult pricing management issues. European countries have agro processes that are far more expensive than countries in other continents. Small- scale farming, high ground prices and high labour costs are factors that influence competition within agriculture worldwide. The European Union helps the European farmer with export subsidies and import levies to enable continuous farming activities in Europe. These subsidy activities of are not accepted by farmers outside the European Union and the (export) subsidies are therefore reduced over time in the future. To enhance fair competition, the European Union is aiming at new legislation ¹⁴ ; where the focus is changing from production subsidies to field/area subsidies. This fair competition driver is a great enabler for agricultural innovation within the European Union. Also trends in the Dutch Agro sector are bases (partly) on this driver. The growing trends of multifunctional farming (i.e. total farming concepts and farming experience) is not only based upon trends within consumer developments, but also needed to enhance agro business performance by using other niche markets. The fair competition driver is important in short term management. A start has already been made with reducing the subsidies for agricultural businesses within the European Union. The fair competition drivers therefore is a short term enabler of agricultural

14

http://www.europarl.europa.eu/news/expert/default_nl.htm

innovation projects. Fair competition is a difficult issue and developments are affecting agro trends.

Fair Trade

An extension of fair competition is fair trade ¹⁵ . The concept of fair trade is concerning honest trade and fair trading prices, especially for agricultural businesses in developing countries. The fair trade driver is not only concerning the agro sector, but also for example textile businesses, fashion industry and lifestyle. Fair trade is part of a social responsibility trend that has started the last years on a relative small scale. In the current situation, fair competition and fair trade are two totally different issues.

However, both aspects are about fair agriculture. Fair competition is only on a smaller scale than fair trade. It will follow the fair competition driver on the mid and long term.

Sustainability

As Prahalad et al. (2009) states, sustainability is the key driver of innovation. In their research, they have determined five stages in developing a sustainable organisation.

However, they also place remarks, especially regarding differences between western society (where sustainability is already started to find its way in legislation and company policies) and developing countries (where environmental issues are less important than obtaining advantage in global competition). This also implies that, in order to achieve global sustainability concepts, fair competition aspects are important.

Sustainability contains issues as bio diversity, climate neutral, recovery of ecosystems and recycling. These issues are in fact al environmental issues. But the concept of modern sustainability consists out of three pillars (Adams, 2006). In figure 13, the three pillars of sustainability are showed. To develop sustainable concepts and make them a real driver of innovation (as Prahalad et al. state), they have to meet advantages on all three aspects. In sustainable agriculture, practices like biological farming, use of pesticides, soil compaction, bio diversity, pollution and packaging are areas of interest. The development of innovative concepts within sustainable farming

15

http://fairtrade.nl/

needs to fulfil also a social and economic value. It is a massive driver for short and mid term innovation for agricultural trends.

Figure 13: Sustainable development model

Ecological Footprint

The impact of human activity on the planet earth is translated into a ecological footprint. The ecological footprint normally measures the amount of ‘earth’ needed to live. The ecological footprint is 1,8 acres. This is based upon the earth and the amount of inhabitants ¹⁶ . However, in developed countries, people use on average at leas 4 times as much. This means that society consumes more than the planet earth can regenerate. With increasing world population and growing wealth perspectives, this consuming of the planet have to stop. An ecological footprint divided in separate categories (e.g. fishing, forests, crops, built up land and the carbon footprint). The goal of the ecological footprint driver is the balance between consumption and regeneration of earth’s resources. Since it is a total concept of all kinds of smaller pieces, which are becoming separate drivers first, this drivers is acknowledged to be a vision driver.

16

http://www.footprintnetwork.org/en/index.php/GFN/page/footprint_basics_overview/

Global Awareness

The sustainability driver will be very important in short and mid term context. In the future, this will lead to a new concept of global awareness ¹⁷ ; a conceptual understanding based upon an applicable knowledge of global and cultural perspectives. This is an extended sustainability concept and contains more aspects than the three sustainability pillars (i.e. environment, social and economic). Global and cultural understandings, based upon the recognition of diversity, are important factors in global awareness. Since sustainability and reducing the ecological footprint are drivers that are important for especially the mid and long term, global awareness is mainly a future visionary driver.

Climate Change

The climate change is an important factor in our society. Of course, it has great influences on environmental issues as the ecological footprint and sustainability. These issues are discussed above, but the climate change also has direct effects on the world and how agricultural businesses innovations. Climate changes cause heavier rains, longer dry periods and higher temperatures. These three aspects are very important in agriculture. New strategies and innovative solutions are needed to cope with these environmental changes. Climate change adaption is the first stage, followed by climate change mitigation actions are important to reduce the effects of global warming.

Especially in Dutch agriculture context this is important, because of the rising water levels and decreasing soil levels ¹⁸ . Climate change is a short and mid term driver.

Energy Neutral

Energy neutrality is a driver that sets in due to increasing energy prices and the fact that fossil fuel buffers are ending. Alternative energy generation is therefore important. Energy neutral means that the amount of generated alternative energy equals the energy needed for activities (e.g. house holds, production lines and farming activities). Governments already started addressing this issue and set goals for the

17

http://www.cccti.edu/globalaware/gahome.htm

18

http://www.rnw.nl/nederlands/article/nederland-zinkt-maar-geen-reden-voor-paniek

future. On the long term, energy neutral solutions will be more at hand. Nevertheless, on the short and mid term it is a real driver for innovation in the agro sector.

Subsistence

An expansion of the energy neutral aspect, is to be not only ‘neutral’ with energy, but also with other consuming products (e.g. food and water). A possible future vision is total subsistence, meaning that households provide their own energy and food, so that they become completely independent of resources of the earth. This possible future driver implies that the agro sector and the supply chain will change completely and radical new innovation is needed.

5.2 Sector

In this research, the market consists out of the Dutch agro sector. This agro sector is divided in three sub sectors (i.e. arable farming, horticulture and livestock farming) and is also shown in the roadmap format. This distinction is made because of the different trends and performance indicators within the different sub sectors. This means that in a further stage, different agro solutions have to be developed to fit these sub sectors.

In the format, there is also room for showing trends of the whole agro sector.

Therefore, this paragraph is discussing the emerging trends and the performance

within the whole agro sector and the sub sectors, using a the timeline presented in the

roadmap format. The research within the Dutch agro sector is done by using agro

publications and discuss preliminary results with farmers in a informal setting. These

settings are also used to let farmers get acquainted with robotics in general and grow

acceptance among famers for agro robotic solutions in general. The results are

described below.