Commercial Indoor Urban Farming:
Operational Challenges to Gain Competitive Advantage
Master Thesis
by
Abstract
Urban farming is seen as a promising part to feed the constantly growing world population in a sustainable way. However, the market for agriculture is highly competitive and the potential benefits of urban farming must be realized in real-life in order to have a chance in the market. Therefore, this thesis focused on (1) the operational challenges for urban farming companies to gain competitive advantage, and from there, (2) analyzed one of the most important challenge in greater detail.
A qualitative approach was chosen that applied a Delphi study to (1) identify and validate operational challenges and then conducts expert interviews to (2) find possible solutions to the most prominent challenge.
While in total nine challenges were validated by the Delphi study, labor cost was considered as the most important operational challenge and further analyzed.
The general findings show that the system of a farm should be designed in a way to optimize workflow. This suggests that the system should allow the planting and harvesting of plants in batches rather than individually. Furthermore, the results showed that the packaging of products has not only an influence on the packing process itself, but also on the distribution and shelf life activities.
The challenges identified raise awareness for urban farming companies, and the already discussed improvements for labor costs support them to improve performance in this regard. However further research is required to validate the challenges, based on an even broader field of experts and analyze the remaining challenges.
Master Thesis
Commercial Indoor Urban Farming:
Operational Challenges to gain Competitive Advantage 28.01.2019
Felix Guillaume Lionel Coudoux S3454444
15509 words (excluding appendix and references)
MSc. Technology and Operations Management
Supervisor & assessor: Dr. J.A.C. Bokhorst Co-assessor: Dr. D.J. van der Zee
University:
University of Groningen, Faculty of Economics and Business Nettelbosje 2, 9747 AE Groningen
Netherlands
Company:
Infarm – Indoor Urban Farming GmbH Glogauer Str. 6, 10999 Berlin
Acknowledgment
I would like to thank my thesis supervisor Dr. J.A.C. Bokhorst of the FEB at the University of Groningen. He always gave me helpful support when I faced challenges within my research and steered me in the right direction. I would also like to acknowledge Dr. D.J. van der Zee of the FEB at the University of Groningen as the second assessor of this thesis.
Furthermore, my thanks go to the experts from the case study company Infarm who gave me valuable insights and made this research possible. I share your passion for a more sustainable world and admire the enthusiasm you are working on this great idea. My special thank goes to the many colleagues and friends I gained throughout my time at Infarm and to the friends I made in course of my studies, whether it was during my time in Pforzheim or in Groningen. Finally, I must express my deep gratitude to my family and especially my parents who were supporting me my whole life, raised me to be an independent man and always encouraged me in my choices.
Last and foremost I want to dedicate this thesis to Kati, the love of my life. You not only provided me with emotional support but also with professional advice writing this thesis and made me a better person throughout our time together. This accomplishment would not have been possible without you.
Thanks to y’all.
Table of Content
List of Figures ...vi
List of Tables ... vii
List of Abbreviations ... viii
1. Introduction ...1
2. Theoretical Framework ...4
2.1 Urban Farming ...4
2.2 Operations Management Performance ...8
2.2.1 Flexibility ...9 2.2.2 Dependability ... 10 2.2.3 Quality ... 11 2.2.4 Cost ... 12 2.2.5 Speed ... 13 3. Methodology ... 15 3.1 Research Design ... 15
3.2 Sample and Data Collection ... 16
3.3 Data analysis ... 19
3.4 Validity and Reliability ... 19
4. Case Study Company: Infarm ... 20
4.1 About Infarm ... 20
4.2 The Farm System ... 20
4.3 Services / Operations ... 22
5. Results ... 24
5.1 Delphi survey results ... 24
5.1.1 Flexibility ... 25 5.1.2 Dependability ... 26 5.1.3 Quality ... 27 5.1.4 Cost ... 28 5.1.5 Speed ... 30 5.2 Production Processes ... 31
5.3 Production Processes Redesign ... 39
5.4 Validation of Redesign ... 47
6. Discussion ... 49
7. Conclusion and Further Research... 53
List of Figures
Figure 1 Research Focus ...2
Figure 2 Commercial UF types (Benis and Ferrão, 2018a) ...5
Figure 3 Infarm Farming Unit V1.5 (Infarm GmbH, 2018) ... 21
Figure 4 Plant Trays and Spirals (Infarm GmbH, 2018) ... 22
Figure 5 Scissor Lift with Harvest Boxes (Infarm User Manual, 2018) ... 34
Figure 6 Box Ground Trolley (Infarm User Manual, 2018) ... 34
Figure 7 Herbs Tray Harvest (Infarm User Manual, 2018) ... 35
Figure 8 Herbs Tray Cleaning (Infarm User Manual, 2018) ... 35
Figure 9 Packaging Area Workstation (Infarm User Manual, 2018) ... 36
Figure 10 Packaging Area Overview (Infarm User Manual, 2018) ... 37
Figure 11 Growing Tray New Design ... 41
Figure 12 Growing Tray New Process ... 41
Figure 13 Danish Cart and Water Tray (B.V., 2019) ... 42
Figure 14 Paper Clamshell Example 1 (SPI, 2019) ... 46
List of Tables
List of Abbreviations
CEA Controlled Environment Agriculture
CO2 Carbon dioxide
CTO Chief Technology Officer
DC Distribution Center
FAO Food and Agriculture Organization of the United Nations FaaSä Farming as a Service
FU Farming Unit
Sec Second
Sqm Square meter
OM Operations Management
POS Point of Sale
1. Introduction
The field of commercial urban farming (UF) has evolved substantially over the last few years (Game and Primus, 2015; Benis and Ferrão, 2018a). Due to the ever-growing population on earth which is predicted to reach 9.8 billion by 2050 (United Nations, 2017), naturally food security is becoming more challenging. Additionally, the progressive urbanization throughout the world has many effects on urban society (Cohen, 2006), food security being one of them (Matuschke and Kohler, 2014). As urban development progresses, agricultural land is being converted into residential land (Riffat, Powell and Aydin, 2016), reducing the amount of land available for agricultural use.
The concept of UF is to grow and cultivate food in the urban environment which leads to the opportunity to improve food security, the local economy as well as social and environmental aspects of a community (Orsini et al., 2013). The forms of UF differ clearly whereby the largest distinction has to be made between commercial and non-commercial urban farms. Commercial UF often takes advantage of controlled environment agriculture (CEA) (Fogg et al., 1979), which promises to grow high-quality food in a shorter period of time and more resource efficient compared to traditional agriculture.
Current literature on UF focuses on exploring the social impacts and food security but lacks the economic implications for companies. Benis and Ferrão (2018) identify the main challenge in establishing UF practices in terms of scalability, resource efficiency and cost-effectiveness, all being influenced by OM. With OM having such a big influence on the success of a business, companies try to gain a competitive advantage through operational performance.
Thus, the focus of this research is on (1) the development and validation of operational challenges of urban farming based on the five operational performance objectives mentioned above and (2) the analysis of possible solutions for the challenges identified.
RQ1: What are the challenges of urban farming companies from an operational perspective? RQ2: What are possible solutions to master the operational challenges faced by urban farming companies?
Figure 1 shows how the research focus on UF is narrowed down.
Overall, the literature on operational activities and the challenges of UF companies is very limited. Therefore, next to academic and grey literature on UF, also literature from the food and agriculture research area concerning the operations performance objectives were reviewed and findings were transferred to the field of UF.
2. Theoretical Framework
The theoretical framework will provide background information for this research and will consist of two parts. First, UF and its different forms will be discussed to narrow down the focus of this research. In the second part, based on the OM Performance objectives, the operational challenges of UF will be explored.
Challenges which have already been discussed in the literature are the general financial feasibility of UF activities and the energy balance of the whole operation. These challenges and studies were not included in the review that was conducted in this work since the focus is on the operational challenges of UF companies once they are established and running daily operations.
2.1 Urban Farming
The notion ‘Urban Farming’ is being used to describe the practice of cultivating and distributing food in an urban area (Mougeot, 2000). The term ‘Urban Agriculture’ is also used very frequently in literature. Agriculture is understood as the practice of farming and cultivation of food (Brassley and Soffe, 2016). However, in this research “urban farming” is chosen as the expression to use.
The variety of food ranges hereby from vegetables and large-scale crop production to animal husbandry, which is the breeding of animals for meat, fiber, milk or eggs. Horticulture on the other side is only the cultivation of plants for food (Shry and Reiley, 2016). In this research horticulture activities in UF are explored since these reflect the most competitive branch of UF (Orsini et al., 2013).
The focus of this research lies on the operation of vertical farms. The method is to raise plants in a controlled environment by using a soilless cultivation method like hydroponics. When using this method, the roots of the plants are hanging in a water-nutrient solution (dos Santos et al., 2013). In a vertical hydroponic system, water is pumped up to the highest point and flows downwards with gravity reaching each of the artificially vertical stacked plants.
Many authors (Despommier, 2010; Banerjee and Adenaeuer, 2014; Benis and Ferrão, 2018a) associate vertical farming with the operation of an entire building where the levels of the farm correspond to the floors of the building. Nevertheless, commercial UF companies like Plenty, BoweryFarming, Aerofarms or Infarm more often interpret ‘vertical’ as having the levels with plants closely above each other in one structure enabling one person to harvest multiple levels at once.
Figure 2 Commercial UF types (Benis and Ferrão, 2018a)
It is surprising that there has been little research on the economic influences of UF so far since the advantages of UF cover a range of OM topics, which have the potential to positively influence the economic performance. The most important benefits of UF from an operations perspective are discussed in the following.
2.1.1 Higher production
The main reasons for a higher yield per square meter (sqm) are on the one hand the multiplication of agriculturally usable land due to using the same amount of floor area multiple times in height. On the other hand, operating farms under controlled environmental conditions allows the farmer to use optimized production methods in terms of temperature, humidity, light exposure and nutritional supply to maximize the growth of the plants (Banerjee and Adenaeuer, 2014).
Additionally, Despommier (2012) identified further advantages influencing higher production including the production throughout all seasons and protection of suboptimal climate conditions leading to yield loss (Bass and Baskaran, 2003). Also, Utami et al. (2012) showed in their research that the growing technique of vertical farming itself can result in higher yields compared with traditional methods.
Touliatos, Dodd and Mcainsh (2016) showed in their research that the yield of lettuce, based on occupied growing area, is 13.8 times higher using a vertical farming system compared to a horizontal hydroponic system and could even be more improved by using artificial light.
2.1.2 Better quality
2.1.3 Shorter travel distance
According to the Food and Agricultural Organization of the United Nations (FAO), one of the drivers for UF is cutting the supply chain and the ecological footprint (Hoornweg and Munro-Faure, 2008). By decreasing the distance due to the proximity to end-consumers, fewer kilometers are traveled resulting in fewer costs for transportation (economic impact), less CO2 emissions and fresher produce. It also means that plants get harvested in the right stage and do not ripen during transportation which results in higher quality.
2.2 Operations Management Performance
Next to marketing, finance and human resources, OM plays a fundamental role in every business. Slack et. al (2016) defines OM as the activity of managing the resources which produce and deliver goods and services. Every company has operations functions because every company produces goods, services or a mix of both. A competitive advantage exists when a company is capable of delivering the same benefits as its competitors, but at lower costs (cost advantage) or delivering benefits to customers that outperform those of competing products (differentiation advantage) (Barney, 2002). Wen-Cheng et al. (2011) recognize internal sources of a firm as critical success factors to gain a competitive advantage.
OM transfers the strategy of a company to its day-to-day operations while transforming input into finished goods and services for the customer (Slack and Brandon-Jones, 2018). For most customers, the price, quality, variety and availability of a product are critical buying decisions. All these factors are influenced by OM. For manufacturing companies, De Toni et al. (1992) recognized early on the performance of the operating system as an opportunity for gaining a competitive advantage. The work of Chase, J. Aquilano and Jacobs (2005) explains in more detail how the competitive advantage can be gained through OM. Likewise, Upton et al. (2004) see more in operations activities as only support roles. They show how a well-designed operations function can become a strategic competitive weapon.
This chapter reviews the literature on operational challenges for UF companies based on the OM performance objectives. Therefore, for each objective, an introduction is given and the challenges in the UF context are explored. At a later stage, the results from the literature review are compared to the results from the Delphi research in order to validate or reject the challenge and discuss possible reasons.
2.2.1 Flexibility
Having a flexibility advantage in operations means the ability to handle unexpected situations or giving customers individual treatment (Slack, Brandon-Jones and Johnston, 2016). To have flexibility in operations implies to react immediately to changes in the customer request (operational flexibility) but also to implement potential changes in the long-term (tactical flexibility) (Carlsson, 1989).
An important agreement in a contract between two parties is next to the specification, the amount and the price of the sold good or service. The same is true for food suppliers and supermarkets. The fact that UF is more flexible in the production due to more harvest cycles compared to traditional agriculture (Riemenschneider, 2018) and also shorter delivery distance, should enable them to offer operational flexibility for short-term ordering and delivery service to their customers.
To properly handle the uncertainty in demand, a system is required which needs to enable the customer to send requests to the company as well as a plant production planning which takes uncertainties into consideration. When these conditions are given and operations are executed properly, UF companies should be able to react to change requests even on the delivery day itself.
horticulture companies themselves (Schoorl, Holt and Mayer, 1986) or for the prevention of pests and diseases (Valiuskaite, Rasiukevičiūtė and Duchovskienė, 2016).
The International Journal of Forecasting (2018) is publishing the special “Food and Agriculture Forecasting” in 2019 covering production, consumption, stocks, trade and prices of major field crops. Overall, the difference between traditional agriculture and UF, which has the possibility to produce crops throughout the whole year and much faster, makes it necessary to take a different approach to forecast sales and production.
Generally, flexibility has to be offered in regard to the mix of varieties and the quantity. Furthermore, changes in the ordering habit of a customer have to be implemented in the demand forecasting to ensure tactical flexibility which again strengthens the operational flexibility.
2.2.2 Dependability
Slack et. al (2016) describe the dependability of an operation as the degree of which a product or service meets the agreed-on requirements in terms of quantity, quality and time at the customer. For building a relationship with the customer, dependability plays a crucial role as clients might move to a competitor if services are always late (Ott, 2010). Based on the advertised benefits, UF has the potential to deliver customer satisfaction regarding quantity, quality and time.
Nevertheless, challenges exist. As UF companies use highly technical farms which rely on the use of electricity, a power outtake can have a significant impact on the yield (Maximum Yield, 2017). A system with hydroponic farms is prone to lose plants if they are not constantly supplied with a nutrient solution. The impact of a system breakdown can also affect the quality, as plants partly decay when not constantly supplied with the nutrient solution. If the breakdown of the system is only over a short time span the plants can survive but might be damaged. In terms of quality, another impact is the non-use of any pesticides. This makes plants more vulnerable to pests and diseases. More on this challenge is explained in the operational performance objective 2.2.4 Cost.
the weather. Furthermore, harvest and delivery can take place on the same day since there are short travel distances.
To reach the objective of dependability the challenge is building a reliable system and having a constant electricity supply to ensure the maximum yield in the right quality. The system of hydroponic farming on its own is not a new technology, however, the scale with which it is operated is. This means the single components of the system, as well as the interaction of them, have to be reliable. Additionally, the electric infrastructure of the building in which the farm is operated must meet high demands.
2.2.3 Quality
Quality is about meeting the expectations of a product or service and for customers one of the easiest ways to judge operations (Slack, Brandon-Jones and Johnston, 2016). Superior quality can create a competitive advantage for a company compared to its competitors. Many papers identified a positive relationship between quality management resulting in better product quality, process performance, perceived quality or reduced costs and the general performance of a company (Jaafreh and Al-abedallat, 2012). For this research, the focus is on the quality assurance of all plant-related activities. Hereby, it starts with seeding and planting the seedlings after the germination. The subsequent activities are called post-harvest activities and range from harvesting, packing and storing of the plants at the company’s location to the transportation and marketing on the shelve of the retailer.
Storey (2016) goes on to address the importance of the right temperature for storing different types of crops where lettuce for example needs lower temperatures compared to basil and the overall importance to keep the cold-chain along the whole post-harvest supply chain. As Midmore and Jansen (2003) point out, a lack of cool storage transportation in developing countries has kept the production of crops close to the customer, whereas Kader (2010) emphasizes the general importance of shipping products in temperature controlled vehicles. As hydroponic horticulture is a soilless system, diseases from soil are not a big threat (Bailey and Lazarovits, 2003). On the other hand, the benefits of pesticides, primarily protecting from pests and diseases (Aktar, Sengupta and Chowdhury, 2009), are not given in CEA. This results in the constant risk of contamination of the farm from the outside which needs to be prevented.
Summing up the challenges for having a superior quality as UF promises it needs to start with keeping hygiene conditions in order to keep pests and diseases out of the production facility. Next, the human factor in the handling of plants needs to be minimized through training on how to operate the different activities. Finally, preserving the quality as long as possible by keeping the plants at the right temperature and maintaining this temperature until they are offered to the end consumer is an operational challenge.
2.2.4 Cost
The operational performance objective of costs deals with the economic perspective on the operations. Low costs are a universal desirable attribute for a company but also for the customers as lower costs of production reflect to the customer in form of lower prices.
Labor is one of the biggest cost drivers when operating vertical indoor farming activities (Agrilyst, 2016; Arnold, 2017). During the “Aglanta” conference one of the panelists, whose company has gone bankrupt, Matt Liotta (CEO of PodPonics) even stated labor-cost as the biggest problem of his company (Michael, 2017). As already proven in healthcare (Marjamaa et al., 2008), improving the workflow of operations decreases the operating time and therefore the costs. Michael (2017) explains that many farms ignore ergonomics and are not efficiently designed for “basic farm operations like planting, inspections, maintenance, and harvesting”. Paul Hardej (CEO of FarmedHere) who was also a panelist at the “Aglanta” conference stated that in the future he would plan a vertical farm first and foremost as a “manufacturing and production process” (Michael, 2017).
Compared to traditional agriculture most activities in UF are still manual. The biggest difference lies in the planting and harvest activities. In most hydroponic systems, plants have to be handled individually to put them in or take them out of the system whereas in traditional agriculture machines are used (Sorensen, Bochtis and Kateris, 2018). In addition to the challenge of labor cost, the cost of electricity for operating CEA is also very high.
2.2.5 Speed
In order to get a competitive advantage in terms of speed, a company tries to minimize the time between the request of the customer and the fulfillment of this request (Slack, Brandon-Jones and Johnston, 2016). The time between these two points is taken by the processing and distribution of the request’s information and the post-harvest operations, which include the internal harvest and packing as well as the transportation to the external customer.
As already mentioned in the performance objective Flexibility, the right system for processing the request and information flow to the right people is key for fast execution. The challenge of improving the post-harvest activities is already covered by the performance objective Cost trying to decrease labor cost.
technology trends such as robotics, automation and self-driving vehicles (DHL, 2018), the consolidation of products in general as well as in special consolidation warehouses (Navickas et al., 2016) is a further solution to improve the transportation strategy within a city.
Table 1 summarizes the challenges per performance objective.
Table 1 Challenges for Operational Performance Objectives from a Theoretical Basis
Performance objective Challenges
Flexibility
§ Operational flexibility for the fast handling of short-term customer requests regarding quantity and variety mix § Tactical flexibility by constantly improving the forecasting
method
Dependability
§ Ensuring the delivery of the right products in the right quantity
§ Ensuring the proper quality when the delivery leaves the production facility and during the transportation
§ Ensuring the timely delivery of products
Quality
§ No contamination with pests and diseases of the plant production from the outside
§ Training of employees for quality assurance throughout all post-harvest activities
§ Keeping harvested crops at the right temperature § Maintaining the cold-chain
Cost § Labor costs
§ High electricity costs
Speed § Request handling in terms of information flow
3. Methodology
The following chapter describes the methodology underlying this research. It deals with research design, data collection and analysis as well as the validity and reliability of the study.
3.1 Research Design
In general, research can be divided into quantitative and qualitative research (Saunders, Lewis and Thornhill, 1997). According to Bryman and Bell (2007), quantitative research makes greater use of statistics and numerical analysis, while a qualitative method increasingly includes words instead of numbers. Moreover, according to Wilson (2013), qualitative data analysis is based on a rather explorative nature.
The aim of this study is to give UF companies a better understanding of how they can overcome operational challenges in order to gain a competitive advantage and thus to support an expansion in the UF market. As already mentioned in the introduction, there is a lack of suitable research literature dealing with the topic of UF, especially from an operational point of view. Considering the explorative nature of this research question and the need to expand theory in this relatively new research area, a qualitative approach is chosen applying a Delphi study to (1) identify and validate operational challenges of the UF, and then conducts expert interviews to (2) find possible solutions to the most prominent challenge.
The Delphi study allows creating a broad research environment with a higher number of experts (Craighead et al., 2007), while the subsequent expert interviews contribute to in-depth research that improves knowledge and makes results more robust.
It is important to note that conducting a Delphi study within research topics in OM is not unusual, as Delphi studies are applied in explorative studies of various kinds, such as technology, product planning and quality management (Subramanian and Ramanathan, 2012). The results are shown in Fehler! Verweisquelle konnte nicht gefunden werden. and Fehler! Verweisquelle konnte nicht gefunden werden..
After the Delphi study, two semi-structured interviews were conducted to gain in-depth knowledge of the most prominent challenge. The use of expert interviews is an effective methodology for collecting narrative data, which in turn allows researchers to explore people's opinions in depth (Kvale and Brinkmann, 2012; Alshenqeeti, 2014).
The interviewees are encouraged to "speak in their own voice and express their own thoughts and feelings" (Berg, 2007), which is essential for the search for solutions to the operational challenges of urban agriculture. The interview questions were directed by an interview guide (see Fehler! Verweisquelle konnte nicht gefunden werden.).
3.2 Sample and Data Collection
As indicated in the previous section, Delphi is an inductive and data-driven process, providing a very efficient and effective way to obtain the views of experts on a particular research topic (Dalkey and Helmer, 1962; Akkermans et al., 2003).
Therefore, a wrong selection of participating experts is considered to be the greatest danger to ensure validity in the Delphi study (Hill and Fowles, 1975; Creswell, 2003). In order to identify suitable participants, a standard method from a previous Delphi research was used: (1) Identification of potential experts, (2) Evaluation of the identified experts and (3) Expert recruitment (Von Der Gracht and Darkow, 2010).
to the small number of companies working in the field of UF and the criteria set, only ten experts from different UF companies were admitted. Within the panel of ten experts, four experts were willing to participate in the study (see Table 2). These four experts work for the case study organization.
Table 2 Participants Delphi Study
Participants of the Delphi study
A Founder and Chief Technology Officer (CTO)
B Team Lead of the Process Design Team responsible for defining the standard processes along the whole supply chain and its requirements
C Senior Process Design Engineer with a focus on the scalability of the services in distribution centers of food retailers
D Senior Supply Chain Planner responsible for the planning of all plant related products
The process of Delphi study applied in this study is based on the variant of the RAND Corporation, which is considered the most recognized and accepted version of the Delphi Method (Dalkey, 1967, 1969). The classic Delphi approach goes through several rounds of data collection. At the end of each round, various points of view and results are identified and summarized for the relevant topic. In the following round, the new knowledge will be available and discussed further (Sekayi and Kennedy, 2017). For this study, the decision was made to conduct three Delphi rounds. According to Mitchell (1991), this approach minimizes research fatigue, thus ensuring a higher response rate and validity.
The Delphi study was conducted anonymously using the online form “Google Form”. All Delphi rounds were carried out between November and December 2018. The data were collected anonymously using the following guidelines and procedures.
In the second Delphi round, the responses from round one were summarized. For each objective, the challenges were listed with explanations and the experts were asked to what extent they agree or disagree and to provide arguments for their response. They were also asked how important it is to meet this challenge for business success. For both, a five-level Likert scale (0-5) was used.
The data from the Delphi study was then analyzed to identify the most important challenge and then to focus on this challenge. Personal interviews with experts were therefore conducted. The data from the Delphi study showed that research must continue to focus on labor cost. Therefore, the selection of the experts was made according to the requirement that the experts must have experience in the workflows of hub operations within the case study organization (see Table 3).
Table 3 Participants semi-structured expert interviews
Participants of the semi-structured expert interviews
E General Manager of the plant hub in the city of Berlin which has a capacity of
around 14.000 plants per week
F Team Lead of the Design responsible for designing workflows and technical
realization of these workflows
3.3 Data analysis
The software "Atlas.ti 7.0" was used for the qualitative investigation of the interviews. This software provides an analytical tool in qualitative research to encode results as they occur in data analysis (Zakaria and Zakaria, 2016). In the first step of the coding process, all interviews were manually transcribed and uploaded to the software "Atlas.ti 7.0". In the second step, the categories are defined in a coding guide, anchor examples are selected and coding rules are defined. After a first material run, the references are color-coded. Then the most frequently mentioned contents of the main and subcategories are summarized and edited.
3.4 Validity and Reliability
Various measures have been taken to ensure the validity and reliability of this study. The internal validity of this study is ensured by using the Delphi method, as the answers of an interview are discussed and compared with the answers of the next interview. In addition, the internal validity is ensured by recording the interviews. External validity is ensured by interviewing different experts with different backgrounds in UF, who have strong expertise in process design, supply chain management, general management and workflow design. Each operational challenge is anonymously discussed between four experts in order to critically discuss and evaluate it. In addition, the challenge considered most important by the experts from the Delphi study was re-tested by discussing it in a semi-structured interview with two other experts.
4. Case Study Company: Infarm
In order to understand the result of this research, it is important to introduce the company with which the research was conducted. The most important aspect is the differentiation between the two types of services Infarm provides: field operation in supermarkets and hub operations respectively distribution center (DC) operations. The focus of this research lies on the latter to get generalizable findings.
4.1 About Infarm
Infarm was founded 2013 by Osnat and the two brothers Erez and Guy. The company produces indoor vertical farming systems. The vision of Infarm is to provide local, fresh, high quality and affordable vegetables to customers in places to which products normally have traveled far distances causing a significant CO2 footprint. The intelligent, modular farms are inspired by nature. The natural processes are reproduced by adapting light spectrum, temperature and nutrients precisely to the needs of the plants in order to maximize taste and growth.
In 2018, Infarm raised 20 million euro in funding and experienced a major expansion due to which the number of employees increased from 40 to over 200. Also, the production output of farms increased by 1000 % and the serviced cities went up from one to in total eight cities at the end of the year.
4.2 The Farm System
In the latest version of the FU, farms exist as a four-level-system and a seven-level-system. This means they can hold in total eight respectively fourteen trays in total. The four-level-system is only used at customers location in supermarkets while the seven-level-four-level-system, which is around three meters in height, is installed in hubs and distribution centers.
The whole system is connected to and controlled through the internet. Air temperature, growing recipe, light intensity and duration can be controlled and manipulated via internet connection.
Figure 3 Infarm Farming Unit V1.5 (Infarm GmbH, 2018)
Figure 4 Plant Trays and Spirals (Infarm GmbH, 2018)
4.3 Services / Operations
Infarm offers a service called “Farming as a Service” (FaaSä). Based on this, customers do not directly buy the farm itself, but rather a guaranteed output per month. All material and operations are provided by Infarm and its employees. This business model originates from the vision of the founders having a decentralized farming network around the city where plants are growing exactly where the customer is buying (supermarket) or consuming (restaurants) them. This type of operations can be referred to as field operations. In addition to this, Infarm also operates hubs in a city to provide supermarkets which cannot have a farm itself and also farms inside of big distribution centers of supermarkets. This operation is referred to as hub operation and is the focus of this study.
Generally, a farm is harvested and planted two times per week in order to distribute the output along the week. Since a farm consists of two columns, on each of the two operational days only one of the columns is harvested and planted. The process of planting, harvesting, cleaning and packing is described in 5.2 Production Processes.
4.3.1 Field operation
The field operation takes care of farms standing at the customer’s location in the supermarket where plants are directly growing in front of the end-consumer. Field growers, mostly living nearby, go twice a week to a farm and harvest, clean and plant new plants. They also pack the plants on site and take care of the point-of-sale (POS) stand. They are provided with the required material (seedlings, packaging, cleaning material) by a driver who transports all the material from the hub to the locations the evening before the field grower goes on site.
4.3.2 Hub and DC operation
5. Results
This section will first present the findings of the initial Delphi survey on the challenges for UF companies. The results of the study led to the focus of the conducted interviews, which is the labor costs of the operations related to seeding, harvest and post-harvest activities. To discuss these results, two sections are structured along these different activities. First, activities are described and challenges are identified. Second, in the redesign section, opportunities to improve the activities are provided by the interviewees and in form of recommendations by the author.
5.1 Delphi survey results
This section summarizes the results from two rounds of the Delphi survey. The third round was not conducted since consensus about the challenges was already found in course of the second round. Table 4 provides an overview of all identified challenges and the corresponding score on how much the participants agree on this challenge and how important they consider mastering this challenge to be for the success of an UF company. The comparison with the challenges found in the literature can be found in chapter 6. Discussion.
Table 4 Challenges for Operational Performance Objectives from the Delphi Study
Performance objective Challenge Score Agree Score Importance
Flexibility Change in the quantity of customer demand 4.25 5.00
Change in the variety of customer demand 4.00 4.75
Dependability Right quantity of plants 4.25 5.00
Right quality of plants 4.00 4.75
Quality Contamination with pests and diseases 3.00 4.00
Storage of plants 4.75 4.75
Cost Operating labor 4.75 4.75
Equipment and Electricity 4.75 4.75
5.1.1 Flexibility
The first challenge for the performance objective of being flexible is the change in quantity of customer demand. This might happen at short notice due to a marketing campaign or good sales performance of the product. B states that “customers expect us to adapt to their demand. Therefore, they are often supposed to receive specific quantities, but change their quantities in the last minute”. D adds to this that “some supermarkets want to have different quantities of the same varieties on different days”. This compromises the planning and standard operations.
Although all participants agree that this problem is not unique to UF companies but also for traditional agriculture, they see traditional agriculture in advantage, because “[traditional agriculture farmers] sell to distribution centers [in] much bigger quantity” (participant B). This gives the supermarket itself the chance to redistribute based on internal demand. And A explains:
“Traditional farmers have a buffer as production costs are very low for vegetables (low energy, [low] rent and high level of automation). [UF] farmers have none of that and the
buffer is much smaller. Being flexible to the market demands is going to make a huge difference.”
D has a similar point of view:
“Supermarkets have many options to buy the same products, so it is very important that UF [is] able to adapt to customer requirements. Otherwise, customer (supermarkets) may go
back to the traditional way what they are doing since many days and what they are comfortable with.”
And although traditional agriculture has again the same advantages as in the challenge before, C sees the positive point that UF can implement the variety changes in the real-life system much faster since it can grow faster while traditional agriculture first and foremost relies on their buffer.
5.1.2 Dependability
Since delivering on time as an issue for dependability is a challenge for every type of company, the focus here is set on producing the right quantity of plants and the right quality of plants.
As hydroponic vertical farming relies a lot on the technology used, downtimes can easily lead to a crop shortfall. Although B states that “this should not be a challenge”, A counters that “downtime is a huge issue and predictive maintenance tools, robust systems and redundancy are needed to ensure customer demands are met”. Nevertheless, in theory, and “with reliable technology, the quantities can be forecasted perfectly as [UF is] not affected by the weather or other problems” (participant B) which would erase this challenge. For the moment the challenge is handled differently: “Through backup … customer demand can be fulfilled.” (participant C). From a business point of view, D summarizes the challenge: “Crop shortfall affects productivity and then profitability.”
Even if the right amount of plants were harvested, this does not mean that all plants can be sold. A big advantage of UF companies is generally that they grow without pesticides. However, this can also lead to the plants being infested by harmful organisms. It is also possible that some plants do not grow as expected and do not reach the necessary specifications. B sees this as an important point, but gives to consider that it should not be a problem as they “grow in a closed and controlled environment” while “traditional [agriculture] is more dependent on weather”. C also endorses that “it’s controllable”. In contrast to that, A admits that this challenge is not fully mastered: “As Infarm is growing produce without pesticides, fungicides or any harmful chemicals, the production sometimes faces downtime or quality issues.”
applies to “completely organic farmers. They do face the same issues.” He adds that “any farming company who wants to be successful will need to find its strategies to overcome those issues.” Furthermore, he claims “boosting quality … can have a tremendous impact. The better the plants, the more they are immune to various pests.”
5.1.3 Quality
The challenge of quality is on the one side as described above a challenge of having a dependable system but also of having a sterile and hygiene atmosphere. Otherwise, plants can get contamination with pests and diseases. Surprisingly this challenge turns out to be the most dissent one with an agree-score of only 3 (1, 2, 4, 5).
While B explains the challenge with the words “the fact that we use zero pesticides and grow plants in closed systems also makes them vulnerable to several threats (fungus, insect, plants disease)”, C adds that since “UF companies are using confined spaces, contamination can be controlled”. Despite B gives a low score on agreeing on this challenge, he still evaluates it as important to have high quality because “it is one of our key arguments for selling”.
Contrary to the statement of C, A argues that “sterile conditions do not exist in nature [and] maintaining sterile conditions where humans and plants are, is virtually impossible”. He suggests “a more holistic” view and says that “plants need to have a strong immune system to be able to tackle the day to day "dangers" of the outside world”. He sees contamination from time to time as inevitable and argues that “it's a sign of good health to show fast recovery”. This would also be very important because as D states: “If [the] customer is not pleased with plants quality, he will not buy any more. This would affect customer experience and future sales numbers.”
product”. In the case of Infarm the “products are still living plants, so they need special conditions and care (they need fresh water) [to] survive longer”.
C also adds the point that “temperature can be a challenge” but argues at the same time that “UF [companies] do not want to store the plants for a long duration as [they] should be served as fresh as possible”. He also admits that “this problem also exists for traditional agriculture” but at the same time he has “the feeling products from traditional agriculture are “stronger”.
5.1.4 Cost
The challenges for the performance objective cost got the highest scores on agreement and importance. This seems logic as cutting cost is the first thing a company tries to do in the process of gaining money. Therefore, the cost of operating labor should be minimized as far as possible.
The operations concerning growing plants (planting, harvest, packing, distribution) are very labor intensive for UF companies. Additionally, compared to traditional agriculture, highly qualified employees (engineers, scientists, chemists) have to be integral parts of the company. Generally, A considers overcoming this challenge as “very important”. D claims that “Seeding/Kitting/Harvesting processes are by now too manual [which] affects hugely productivity and then costs”. B also asserts this challenge as true and finds next to the used equipment several elements concerning operating labor costs. He sees that “the technology and development of the company rely on highly skilled people, that are usually expensive labor cost” and also “the time to harvest can be very long due to the business model of the company”. Herewith he means the system and process in which plants are grown. This has a direct impact on the labor cost for operating the farms.
D also adds that “operating cost of the end product varies based on the location (operating cost in South Germany and North Germany is different)” which shows the importance of having standardized, waste-free processes.
„Automation is making its way into indoor farming as more money (& engineers, scientists, chemists) is introduced into the system. In the very near future, a huge drop in labor cost will
be needed and seen. The opportunities for automation are huge, and that is a part of what investors in the industry are looking for.”
Compared to traditional agriculture, B sees them in a cost advantage as “they use bigger machines and less workers to harvest fields” while C also detects an advantage for UF companies as “compared to traditional agriculture other overhead costs are much less (for e.g. pesticides, fertilizers, soil maintenance, water etc.)”.
The second challenge concerning the cost objective deals with the cost of equipment and electricity. The equipment which is necessary to build and operate a vertical farm is expensive as it is very technical. In addition, the amount of electricity that this equipment consumes is very high.
A recognizes the challenge about the electricity usage as “one of the most important topics for an indoor farming company as every percent count”. He sees many efforts in the whole industry to overcome this challenge for example in the field of “smart battery tech[nology], optimizing algorithm for power consumption, more efficient LEDs, precision lightning and irrigation”. However, this challenge is not unique to UF companies which grow vertically. Also “greenhouses are large, highly sophisticated machines using lights, climate control, water management and more” (participant A). B offsets the energy cost with the transportation by saying that “urban farming saves a lot of money in goods transportation, but a part of this saving is then spent on electricity cost”.
For the costs of equipment, A goes into detail on the farm itself and explains as “indoor farming tries to mimic natural conditions using mechanical/software features, indoor farming production machinery is expensive”. This “poses a big challenge for business model design [as] clients are not willing to pay for the machinery itself” (participant A). Nevertheless, he does not see this negative as “this leads to the evolution of new and exciting business models to finance the high capex machines [and] the introduction of venture capitals who are somewhat more aligned with those issues than traditional investment vehicles.”
5.1.5 Speed
Supermarkets rely on getting products as fast as possible to be able to sell them to customers and gain money. This means that the time between request and delivery should be as short as possible. The fact that growing these plants takes around 30 days and supermarkets are normally used to have a requested good ready in store after 24-48 hours, is an additional challenge.
A notes for this challenge that an UF company “needs to offer a higher level of service, in an overall cheaper price. That means faster deliveries, higher quality, higher reliability, and reduced costs”. A solution D offers is “to set-up extra buffer quantity to provide customers really quickly”. B also says “the volume … helps coping with this challenge”. At the same time, he sees the promising trend that “due to the amount of data and the close relationship … to our customers, we understand better the market than regular farming does”. This leads to “getting better at forecasting the demand and knowing what sells best”. C agrees that “this challenge can be solved by good planning and strategies”.
Overall, traditional agriculture faces the same challenge which might be even more critical for them “as they need much more time to grow” (participant B) and “are often far away from the cities” (participant D). However, the arguments (see 5.1.1 Flexibility), that traditional agriculture sells more often to distribution centers and has a bigger buffer cannot be neglected here.
5.2 Production Processes
In the following sections first, the processes are described and challenges are identified. Then, improvement ideas indicated by the interviewees are presented. In addition to this, redesign ideas and improvements by the author are explained and later on verified.
5.2.1 Seeding
The seeding happens inside the nursery. The seeds are being put inside a growing medium. This growing medium comes in rectangular trays with 240 growing medium plugs in it. One tray always contains one variety.
At first, when the whole operation started there was one person “with a small spoon putting [the seeds] into the growing medium” (interviewee F). Today, they use a “semi-automatic seeding machine which is like a machine with some tubes everywhere where you can do it in one shot” (interviewee F). Due to this improvement, the time for this process was reduced drastically.
5.2.2 Germination in the nursery
5.2.3 Kitting of seedlings
When the plants have germinated into seedlings, they are moved into the kitting room. This part is “a very, very labor-intensive job” (interviewee E) as each seedling is taken out of the trays, placed into a plastic pot and then put into a box for further handling.
One reason to do so is to protect the growing medium in which the plant is growing in. The plastic pot protects it from deformation. This deformation might happen while the plant is inside the farm and the spirals are being rotated. Due to the force which is applied here, without “the capsules, it would just snap the seedling” (interviewee F).
A second reason for this activity is to break up the big trays of 240 seedlings for easier handling – physically wise and organizational wise. When kitting seedlings in boxes, one box is dedicated to one specific farm. This box can carry different varieties which makes it easier to control what comes in a farm.
The process itself is organized as a production line. Before it starts, the production order is received from the IT system. This system also generates the stickers which are applied on the boxes carrying the information about variety mix and to which farm it belongs. Based on this information, the first station prepares the right number of boxes and places metal grids and plastic capsules. One box can hold up to 77 seedlings. The second station then “take[s] the seedling trays, set[s] them in the center of the table and … transfer[s] the seedlings from the seedling tray to the capsules in the grid” (interviewee F).
By performing the task of seedling kitting like this, F argues that the “process gets easier and more understandable” when preparing all the grids with capsules first and then filling them up in a separated step. Furthermore „batch[ing] the actions … is actually faster to do … rather than repeating again and again the same process” and F states additionally that less space is necessary when working in a production line since the material for a certain activity is only needed at one station.
5.2.4 Planting
In an up and running system, the planting takes place after the harvesting and cleaning in order to have space for the new seedlings. In this activity, the seedlings, which are now bedded in a capsule are taken out of the kitted boxes and put into the farm. “Planting [takes] around, depending on the variety, around 40 seconds to one minute per tray” interviewee F analyzes and goes on to describe that for level one to four the grower is “using a step ladder” and “a scissor lift from the tray five to seven”.
5.2.5 Harvest and cleaning
Just like the planting, the harvesting and cleaning activity is completely manual and designed to be split up between two workers. One is working from the ground level and using a step-ladder to operate level one to four and the other one is using a scissor lift to operate level five to seven.
For the work content of the two workers and the time needed to perform the tasks, F states: “It’s kind of evened out because the one on the scissor lift has to drive around not too much. You can count a minute to go from one farm to the next one, but then the other one for the two
first level, he can do it from the ground where it is extremely fast and for the third and fourth level he can use the step ladder which is a bit less effective. While the one on the scissor lift
has to adjust his position.”
Figure 5 Scissor Lift with Harvest Boxes (Infarm User Manual, 2018)
Figure 6 Box Ground Trolley (Infarm User Manual, 2018)
Figure 7 depicts the process of harvesting a full tray of herbs. First, the worker takes the plants which are in the front half of the tray and places them inside the harvest box.
Afterwards, the spirals have to be turned clockwise in order to bring the back side to the front. It is important to rotate clockwise to make sure the spirals are not opening up.
Figure 7 Herbs Tray Harvest (Infarm User Manual, 2018)
After harvesting a full tray, the worker has to clean it to prepare it for the new seedlings. Therefore, first the outer tray is cleaned with a cloth. Afterwards, the spirals have to be turned counter-clockwise which brings new plants to the outer tray and opens up space in the inside to plant new seedlings. This opened up space in the middle needs to be cleaned as well.
Figure 8 Herbs Tray Cleaning (Infarm User Manual, 2018)
5.2.6 Packing
Each DC block has a packing area where two workers can work simultaneously. In Figure 9 the two workstations can be seen in a frontal view. This specific overview shows how a station looks like when herbs are not packed as singles but as a set of three for the culinary business. In this case, a sticker dispenser is needed to close the packaging. For the normal packaging, no sticker is needed.
Figure 9 Packaging Area Workstation (Infarm User Manual, 2018)
Figure 10 shows an overview from the top of the packaging area. Again, the two workstations can be seen. What is not shown in this overview is that each packing area needs to have a sink due to hygiene issues. Behind the workers, there are two trolleys standing. One is full of boxes with harvested plants and the other has the same number of empty boxes.
Interviewee E explains the setup and purpose of the area as follows:
“The goal is for them to have the access to all and each of every item which they need to make the packing. For this [reason], they will do the packaging in bulk which means they will
Figure 10 Packaging Area Overview (Infarm User Manual, 2018)
The process starts for each worker by taking a full harvest box and an empty box. They place the full box on the left side of their workstation and the empty one either on the right side on the workstation or on a box ground trolley as seen in Figure 6. F explains that they then “pack it and fill up those boxes which then go back into the packed goods cart”.
Interviewee E explains the process more detailed:
“You need to take [the plant], you need to wrap the root and then we put this bamboo skewer on the roots, so the roots do not hang loose. And then you need to take the packaging itself, which is a paper bag, put it in and put it in a box. Per plant, you need at least five seconds or
maybe more. Depending on how trained the grower is.”
In Fehler! Verweisquelle konnte nicht gefunden werden. and Fehler! Verweisquelle konnte nicht gefunden werden. the complete and detailed operating procedure can be reviewed.
5.2.7 Storage - Kitting of plants – Distribution - Shelf-life
These activities are grouped together in one section. After the harvested plants are packed, they are brought to the storage where they are waiting to be kitted to order. It can also happen that orders are already kitted before and then these kitted boxes are brought to the storage. In the storage area, it is important to keep the right environment. With “light, … water and … a temperature-controlled environment, the plants stay fresh for one week or even 10 days” E clarifies.
The distribution from a hub to the supermarkets takes place in a temperature-controlled vehicle. For DC activities the distribution will be taken care of by the supermarket.
5.3 Production Processes Redesign
In this section, the processes described before are being redesigned. For each step, points are discussed which can be improved. On the one side, the interviewees indicated points where they see a potential for improvements and on the other side the author analyzes these points and shows also areas which can be further improved. This redesign aims for improvements of the production processes being more efficient and hence reducing the cost related to labor and equipment. To achieve excellence in these processes carries a big potential to achieve a competitive advantage over competitors and increase the revenue margin.
5.3.1 Seeding
The seeding already experienced a major improvement by changing to the “semi-automatic seeding machine” (interviewee F) compared to the manual method. Recently Infarm bought a “seeding machine which is a solution to reduce time” (interviewee E). With this machine up to 600 trays an hour can be seeded. This means this machine has the capacity to supply easily all farms of Infarm with seeded trays. Also, this machine is suitable for further expansion.
5.3.2 Germination in the nursery
The germination of the seedlings is not a labor-intensive activity. Nevertheless, the nursery itself where the germination takes place is very costly and takes up a lot of space. F mentions that „the main idea is to move completely to nursery farms, just in terms of [the] footprint of the nursery”. With this step, it is not only possible to save space in a new building, but also to save the time building a nursery resulting in having a running business faster. F comments on this point that “a nursery farm means [that] it is installed like a normal farm. So, within a day you can have an up and running nursery”.
5.3.3 Kitting of Seedlings and Planting
The kitting of seedlings, as well as the planting are activities which are either being optimized based on the given circumstances (kitting) or so simple that improvements are not possible (planting).
However, the whole process of kitting and planting is something which should become obsolete. As F explains “the final aim is really: You seed the plants, you put them in the machine and then … the plants come back fully grown. That’s the target we are really aiming for”. While this seems like the best-case scenario, there is one important point which needs to be considered. While right now one growing medium tray has approximately the same size as a tray with spirals, it holds 240 growing mediums compared to 60 plants in a plants tray. Even though the spirals are not space-efficient due to their round shape in a rectangular tray, plants need a certain space to grow. Also, it is also easier to harvest when plants are not packed close together. This means the trays with growing mediums need to be designed in a way that plants have enough space and fit in the irrigation system of the farm. This leads to more effort and therefore costs for the supplier of the growing medium on the one side and on the other side to less efficient space utilization in the growing medium trays. Hence, more air is transported and more growing medium trays are needed. All of these risks have to be taken into consideration and a cost analysis has to be made.
Even if it turns out to not be cost-efficient to do so, there is another point which can be improved. The process of kitting for farms inside the hub is “something [Infarm] would like to remove at one point … but for now, it’s necessary” (interviewee F). The reason is that, as described in section 5.2.3 Kitting when turning the spirals, they would crush the growing medium without a capsule. A solution can be to move away from spirals either to the system where the growing medium trays are used directly or to linear trays. The latter option is currently already investigated by Infarm. Therefore, a focus should be to design the grids in a way that a seedling can be placed in it without a capsule to hold it.
prepared grids with seedlings are placed in the back of the tray. In Figure 11, a sketch of the design is seen and in Figure 12 the proposed process is shown.
Figure 11 Growing Tray New Design
Figure 12 Growing Tray New Process
5.3.4 Harvest
Due to the activity described above not only the time and effort for the planting, but also for the harvesting can be drastically reduced since instead of taking twenty single plants and placing twenty single seedlings, only two movements are necessary for each.
Additionally, by not having to turn the spirals time can be saved for the grower as well as the cleaning will get easier since only one layer (spirals are two layers on top of each other) has to be cleaned.
The dimensions as well as dividing one stage into two trays have a particular purpose for the harvesting. For the transportation, Infarm uses so-called ‘Danish carts’ (see Figure 13) as they are affordable and modular in terms of how many levels can be assembled on them. In combination with a water tray, Infarm can stop using harvest boxes.
Figure 13 Danish Cart and Water Tray (B.V., 2019)
42 % of the original value which might also have an influence on the layout of the packaging area since more carts would be needed to store the same number of plants. Whether a layout change is actually necessary and feasible, would go into too much detail at this point but needs to be studied.
Another point that F states in the interview is that “the main improvement from the operation standpoint is … the verticality of the farm”. This leads to a separation of workflows and the necessity of scissor lifts. This is a major challenge, which has its origin in an actual advantage of vertical farming. A potential solution as Paul Hardej already stated is to design a vertical farm first and foremost as a “manufacturing and production process” (Michael, 2017). As a result, trays with plants could be only as high as one person can reach without any equipment or only a simple step ladder. Usually, industrial buildings, as the ones UF companies operate in, have a huge volume and not necessarily a big ground area. To exploit the characteristics of the building and realize the potential of the verticality, mezzanines are a solution. A mezzanine is an extra floor between the ground floor of a building and the next floor up or the roof. Thus, it gets possible to operate a high farm without having to use scissor lifts. Nevertheless, it might still be necessary to use a step to harvest the upper plants.
5.3.5 Packing
The process of packing the plants is very important as it determines how they are presented to the customer. Thus, training on how to properly wrap the roots and place the plant in the packaging is essential. Packing one plant takes around five seconds. This time varies “depending on how trained the grower is” (interviewee E). While this might not seem to be a lot, it turns into an issue regarding the volume which has to be packed. Thus, it becomes evident that training is not only necessary for quality assurance but also for the reduction of the packaging time.
