Keywords

1 Introduction

Adopting new technology and AI-driven techniques may have the potential to, i.e., optimize production, increase insights and gain a competitive edge thus leading to economic growth for companies as well as making them more sustainable [1]. This paper addresses the willingness, ability, and plans to adopt digital technologies, or AI (artificial intelligence) in a rural region of western Norway called “Sogn og Fjordane”. We wanted to find out to what extent SMEs in the region use data analysis and AI applications and the companies understanding and skills of how AI and data utilization can change their businesses. We also wanted to find out if they had plans for technology development or further data utilization.

The small population density of Sogn og Fjordane (5.9 people per km2) makes it a rural region. The business life is dominated by small-medium sized enterprises (SMEs) [2]. However, despite being a rural region reliant on SMEs, it is also renowned for its robotics development through some companies and university groups, and several nationally and internationally known digital companies and companies known for being at the forefront of technology have their primary location there. Whether these forefront companies and other activities have affected the general business life in terms of digitalization, AI, and technology development is unknown, and another interesting research question to assess.

The contribution of this paper is to uncover the digital maturity state, including AI usage among SMEs in a region, and, from a regional perspective, discuss some potential key future aspects and opportunities for sustainable data and AI adaptation.

2 Literature Review

2.1 Industry 4.0 and Value of AI in Future Business Life

There are numerous definitions of AI depending on the targeted usage of it. The European AI Strategy defines AI as: “refers to any machine or algorithm that is capable of observing its environment, learning, and based on the knowledge and experience gained, taking intelligent action or proposing decisions.” [3]. Like humans’ ability to learn from past experiences, AI can learn from historical data. By analyzing data, AI-powered systems can detect patterns and make informed decisions that lead to optimal outcomes. These outcomes can range from increased efficiency and process optimization to the prediction of future outcomes, providing decision support, and automating traditional manual and repetitive work.

The fourth industrial revolution (Industry 4.0), or “the digital shift” [4] is upon us and over the last two decades many businesses have undergone a drastic digital shift, starting with less paperwork and more digital data to AI driven business models. As the amount of data captured, collected, and used globally along with AI-applications continues to increase rapidly [5], exploiting data and such technology is becoming an increasingly important asset for governments and enterprises. In efforts of meeting global standards on quality, technology, sustainability, and pricing, leveraging data and Artificial Intelligence (AI) seems to be a continued growing key ingredient of Industry 4.0 [6]. Some studies suggest that AI-technologies may contribute to boost a company’s revenue by 10–70% [7, 8]. By 2030, projections made by PwC suggest that AI-technologies (including robotics) has a potential to create 15.7 trillion dollars in annual value globally, making it the largest commercial opportunity in today’s changing economy [1].

The current focus on circular economy and sustainability also creates a huge pressure on making manufacturing operations more ethical and sustainable. The application of Industry 4.0 technologies has been observed to have positive influences on achieving those goals [9].

2.2 SMEs and Challenges of AI Adoption

In the race to adopt new digitalized methods, sustainability, and a circular economy, some companies are at the forefront and others are lagging [1]. A larger study of AI usage in Nordic countries based on interviews conducted in 2021 found that 70% of Nordic companies’ overall use of AI as part of their product and 55% were planning to experiment with AI techniques within the next 6 months [12]. However, the threshold of market entry is rising, and business development is dominated by large enterprises. In Norway, only 9% of small enterprises used at least one AI technology, while the adoption rate was twice as high for medium-sized enterprises. For large enterprises (250 + employees), 43% used more than one AI application. Roughly 76% of companies answered that they have plans to hire new AI professionals within the next 6 months.

Small and medium enterprises (SMEs) often play a crucial economic role in many countries, especially in less urban areas, as they stimulate innovation, provide jobs, foster competitiveness, and contribute to overall economic growth [10]. Several studies have assessed and reviewed why SMEs face various challenges in adopting emerging technologies such as artificial intelligence (AI) [e.g., 11, 13]. Here, key challenges in SMEs are found to be related to a lack of necessary strategies and knowledge (often due to a lack of knowledge among leaders), lack of talents and skills, and a lack of resources. The latter is often found to be related to economic shortage or being unable to obtain large amounts of high-quality data and use cases in small companies [11, 13].

3 Methodology and Empirical Data

A 6-year-long research project, “Teknoløft Sogn og Fjordane”, was financed to build competence in digitalization and automation, robotics, and big data at the Western Norway University of Applied Sciences (HVL) and the Western Norway Research Institute (Vestlandsforsking) and conduct research in close cooperation with local enterprises in the region of Sogn og Fjordane. The overall target of the project is to increase the role of research in local industry innovation.

Sogn og Fjordane has a population of 109 000 spread over an area of 18 623 km2 [14]. This makes it one of the most rural regions in Norway in terms of population. According to Statistics Norway, 99% of companies in Sogn og Fjordane have less than 49 employees, and 90% of companies have 9 or fewer employees [15] (Fig. 1). Agriculture, forestry, aquaculture, and domestic trade occupy the largest sector among small companies (0–20 employees). The latter sector has an annual turnover of almost 14 billion (as of July 26, 2021), which is ~30% of Sogn og Fjordane’s total annual turnover. Some larger businesses with more than 100 employees are in education, human health and social work, manufacturing, transportation, and storage, in addition to maritime industries. The region is abundant in natural resources, which has enabled the development of industries within agriculture, aquaculture, hydroelectricity, aluminium, and petroleum production. Several of the oil and gas fields that contribute to the country’s welfare are located right outside the coastline of the region. The coastal city of Florø is, for example, Norway’s largest supply base for the petroleum industry [16]. The landscape of the region is otherwise largely characterized by lush green fields, mountains, fjords, and glaciers, the most notable of which is the Sognefjorden, the longest fjord in Norway and the second longest in the world. The area has seen an influx of people and businesses due to the trend of younger generations seeking a more rural lifestyle close to nature. Analytics suggest that the region is expected to experience an increase in population over the next two decades, contradictory to many other rural regions in Norway [17].

Fig. 1.
figure 1

Evolution of employee size in companies in SFJ between 2009 and 2019 (pre-COVID evolution) 2019 was also a year when Sogn og Fjordane became a part of the larger county of Vestland. As seen, the number of active micro-companies has grown significantly during this period.

Full size image

3.1 Survey

To map the status of digital maturity and AI adoption among enterprises based in Sogn og Fjordane, we collected answers through a questionnaire. The first part of the questionnaire entails general information about the company and the work position of the respondent. The main part presents many allegations essentially about the company’s thoughts and plans on digitalization and AI, while in the last voluntary part, the respondent can explain with their own words their plans, challenges, and competence needs regarding digitalization. The last part is an addition to the otherwise identical questionnaire sent out to businesses in other nearby regions.

We got answers from 25 respondents from various companies across a range of sectors (Fig. 2). However, three companies that answered are based outside the region and therefore excluded from the survey responses.

An almost identical questionnaire, “Digital maturity in western Norway” was collected by Bergen Næringsråd and others in 2021 (BN survey), where they got ~350 replies [18]. However, even though the region of “Sogn og Fjordane” is a significant part of western Norway, no businesses from that region were represented in that survey [18]. Hence, when collecting answers from this missed region, the similarity of the questionnaires was important to enable a future comparison and a merge of the answers from the greater region.

Fig. 2.
figure 2

(A) Employee-size and (B) sectors of enterprises that joined the survey

Full size image

4 Hurdles of Data Leverage and AI Implementation Among Regional SMEs

Figures 3 and 4 present the survey results. 10 of 21 respondents answered positively regarding their company’s understanding of how AI, machine learning, or robotics can affect their business development. Furthermore, 11 agrees that their company is aware of how they could change or improve with data utilization, and all of those agreeing to that statement also state that their company has concrete plans or strategies related to exploiting their data. Five companies answered that they are lacking concrete plans to leverage their data, and all of those also disagreed that their companies understood the value of AI from an enterprise perspective. Nine respondents answered that AI is complicated and that it is hard to know where to start with it. When comparing small-medium-sized and large enterprises regarding future and AI understanding, the answers are comparable, meaning that we cannot identify a trend of less or more digital maturation depending on the company size based on the respondents’ answers (Fig. 3).

7 of the respondents expressed apprehension that the key employees may not have enough understanding about the significance of upcoming technological challenges, and 6 confirm that there is limited or no financial framework assigned for digital development within their companies. Furthermore, 10 companies answered that they only prioritize simple digitalization work (Fig. 3).

A noticeable inclination was towards inadequate regular competence training to enhance employees’ digital competency, with 11 companies agreeing to this statement. Maybe a significant point, as 8 expressed difficulty themselves in comprehending AI and machine learning, feeling that it is a complex area to initiate with. Five of the respondents were neutral to that question. In addition, five respondents anticipate that AI and robots may become their biggest rivals in the future.

Despite these findings, 16 of the respondents reported that their company has definite plans to tackle future technological challenges (Fig. 4). The three that disagree that their company has plans for tackling future technological challenges partly or strongly agree that they are worried if enough of the key employees in their company understand the seriousness of future technological challenges.

As much as 6 of 7 leaders and managers agree that their company understands how AI and/or robotics can change their business, while significantly less than 4 of 12 regular employees agree with the same statement (Fig. 4). The leaders that agree state that their company also understands the value they can create with their data and has plans or strategies to exploit their data better in the near future. Interestingly, 3 of these leaders also agree that AI is an area that mainly concerns the IT department or IT-responsible and 2 of the respondents who agree that the company understands how AI and/or robotics can change their business also agree that AI is complicated and it’s hard to know where to start.

The ones disagreeing that their company has plans or strategies regarding the exploitation of data in the future (6 respondents) are either not aware of the value they can create with their data (5) or neutral to that (1), and 5 of 6 are worried if enough of the key employees in their company understand the seriousness of future technological challenges.

To conclude, it seems that a significant portion of the respondents, especially leaders, believe that their companies have a good understanding of how AI and data utilization can impact their business development, and these are companies that also have plans for the exploitation of their data and are facing future technological challenges. Some companies lack concrete plans or strategies to exploit their data, and those are also the companies that express apprehension or limited understanding of upcoming technological challenges, including AI, and would only prioritize cost-efficient and simple digitalization work. There is an inclination towards inadequate regular competence training to enhance employees’ digital competency, and there seems to be a discrepancy between leaders, managers, and employees in terms of their perception of the company’s understanding of AI and robotics, with leaders being more positive.

Fig. 3.
figure 3

Allegation answers from the survey on digital maturity - positive skewed answers

Full size image
Fig. 4.
figure 4

Allegation answers from the survey on digital maturity - negative skewed answers.

Full size image

5 Usage of AI in Regional Key Sectors

Here, we will present some examples of AI usage in key SMEs—sectors of the region—to shed some light on the possibilities within these sectors regarding data utilization and AI.

5.1 AI Potential in Agriculture

The experience of AI implementation in agriculture and aquaculture worldwide has shown great potential for higher value creation as well as often affecting sustainability positively [19]. The agriculture sector is increasingly collecting data that can be used for AI applications, or so-called precision agriculture. Sensor data, images, and satellite images are used to classify and detect different phenomena and objects, as well as for performing sorting procedures. Some examples of the monitored objects include weed and disease detection, yield predictions, soil and water management, species recognition, crop quality, animal welfare, and livestock production [20].

A specific target known as one of the main challenges in agriculture’s crop production is weeds and diseases [20]. Accurate weed and disease detection is necessary for sustainable agriculture because weeds are competing for the same resources as crops in terms of nutrients in the soil and sunlight. Today, pesticides are one of the primary methods for combating weeds and diseases; however, pesticides contaminate soil and drinking water, especially in Norway, where the low temperature makes the breakdown of pesticides slower [21]. Machine learning algorithms, in conjunction with sensors and images, can enable the detection and identification of weeds and diseases without causing environmental issues or secondary effects [20]. This may be executed with cameras mounted on drones, four-wheels, or tractors in the field, which automatically take pictures of the same location and of the same objects in the field.

Moreover, image processing and analysis of the pictures may involve counting fruits, classification of colors, and eventually appearance or changes in fine-scaled objects on the fruits, such as diseases. While some manufacturers invest in AI-based sorting machines, most sorting today is still done manually [13]. AI-based apps integrated with machines have the potential to be more efficient, less biased, and more accurate, in addition to replacing boring, repetitive work at workplaces [22, 23]. At fruit and greens reception sites, fresh food turnover time, product quality, and waste reduction are essential. An AI-based sorting machine can use automated image recognition for quality checks and sorting and significantly speed up the sorting process, as well as making it more accurate. The algorithms can be trained to identify flaws, contaminants, or product defects [23]. In Sogn og Fjordane, several such sorting machines are already placed at the region’s largest fruit and greens reception site. Experiences of AI implementation in agriculture and aquaculture worldwide has shown great potential for higher value creation as well as often affected sustainability positively [19]. The agriculture sector is increasingly collecting data that can be used for AI applications, or so-called precision agriculture. Sensor data, images, and satellite images are used to classify and detecting different phenomena and objects, as well as for performing sorting procedures. Some examples of the monitored objects include weed and disease detection, yield predictions, soil and water management, species recognition, crop quality, and animal welfare and livestock production [20].

A specific target known as one of the main challenges in agriculture of crop production, is weeds and diseases [20]. Accurate weed and disease detection is necessary for sustainable agriculture because the weeds are competing for the same resources as crops in terms of nutrients in the soil and sunlight. Today pesticides are one of the primary methods for combating weeds and diseases, however pesticides contaminate soil and drinking water, especially in Norway as the low temperature makes the breakdown of pesticides slower [21]. Machine learning algorithms in conjunction with sensors and images can enable detection and identification of weeds and diseases without causing environmental issues or secondary effects [20]. This may be executed with cameras mounted on drones, four-wheels or tractors in the field which automatically takes pictures on the same location and of the same objects in the field.

Moreover, image processing and analysis of the pictures may involve counting fruits, classification of colors and eventually appearance or changes in fine-scaled objects on the fruits such as diseases. While some manufactures invest in AI- based sorting machines most sorting today is still done manually [13]. AI-based apps integrated with machines have potential to work more efficient, less biased more accurate, in addition to replace boring repetitive work at workplaces [22, 23]. At fruit and greens reception sites, fresh food turnover time, product quality and waste reduction are essential. An AI-based sorting machine can use automated image recognition for quality checks and sorting and significantly speed up the sorting process as well as making it more accurate. The algorithms can be trained to identify flaws, contaminants, or product defects [23]. In Sogn og Fjordane, several such sorting machines are already placed at the region’s largest fruit and greens reception site.

5.2 AI Potential in Aquaculture

One of the largest industry sectors in Norway is the aquaculture sector [24]. An already-published report presents an overview of the most common use of big data in aquaculture and comes up with recommendations that can be applied to businesses within the Sogn og Fjordane region. The report also highlights areas of research needs that could potentially be converted into research projects for businesses in the region [25]. These recommendations are, i.e., water quality combined with fish behavior monitoring using sensors and underwater cameras, techniques to make fish feeding more efficient and reduce feed waste (identified as one of the largest challenges in aquaculture), and disease identification using underwater cameras and biosensors. It has been shown to contribute to improved operations, leading to faster production rates, better asset control and prediction, and enhanced safety and efficiency. Due to the implementation of data-driven computing, it is possible to predict sea lice two weeks in advance.

Additionally, enormous quantities of data are also currently either in the planning phase or being collected related to a global push to track small-scale fishing farms and vessels [25]. These quantities of data are of great variety, ranging from fishways, fishing boat administration, marine satellite images, readouts of fish farms, tides, weather, etc. While this data is increasing and becoming available as sensors and tracking become more widespread, companies are also concerned with managing the substantial complexity of their data [25].

5.3 AI Potential in the Hydroelectric Energy Sector

Like aquaculture, the energy sector is experiencing a similar increase in sensory data that you also see in aquaculture. Continuous measurements of grid equipment and related parameters can be used in predictive analyses for developing maintenance strategies [26]. The most common and well-known source is the smart meter. Prediction of outages in the power grid is essential to having a reliable power grid. Predicting the influence of weather, smart meter events, and location will help identify root causes. For example, researchers at GE Research used GIS (geographic information systems) data and satellite imagery together with survival models to assess outage risk [27]. This was used to find outages related to vegetation. It has been found that vegetation events are not only due to growth but also to extreme weather events. Integrating weather forecasts into the systems for outage prediction is necessary for even better predictions. Other data sources more relevant to the region, such as SFJ, may also be a point of research in the future. Another approach may be to use machine learning and Bayes decision theory to find the optimal decision boundary. The proposed classifier provides an effective framework that not only minimizes outage prediction errors for power system components but also considers the cost of each preventive action according to its implication in extreme events [28].

Smart meters are also installed at the end-user’s location and track energy consumption with a resolution as high as minutes or even seconds [29]. The smart electricity grid enables a two-way flow of power and data between suppliers and consumers to facilitate power flow optimization in terms of economic efficiency, reliability, and sustainability. This infrastructure permits consumers and micro-energy producers to take a more active role in the electricity market and dynamic energy management. Robust data analytics, high-performance computing, efficient data network management, and cloud computing techniques are critical to the optimized operation of smart grids [30]. Weather data, thermostats, real estate data, and energy behavior integrated with energy demands can provide better forecasting and prediction services [26]. GIS also has an important role in the sector. Data from GIS sources can provide valuable information that can be used in decision-making systems since it provides local geographic information for many issues, such as identifying optimal locations of solar farms [26].

Furthermore, forecasting the power generation potential, or streamflow to power plants, forward in time using AI-based time-series methods and weather forecasts has shown great accuracy and uses, providing a potentially powerful new tool for renewable power companies that have not already implemented the method [31].

5.4 AI Potential in Manufacturing

A food manufacturer company called Danone Group started to use AI to improve their demand forecast accuracy, which they report has led to a 30% decrease in lost sales and a 50% reduction in the demand planner’s workload [32]. Manufacturing, which is also large in Sogn og Fjordane, is another interesting sector as it has great potentials regarding circular economy, sustainability, and otherwise high AI potentials [33]. By using AI, businesses can react and have better control of market responses, identify trends in demands and consumer behaviors, conduct forecasts, and automate complex tasks and decision support, therefore maybe freeing up employees to more interesting innovation tasks [34].

Manufacturing also involves product lifecycle management, operation and maintenance, energy management, and supply chain management. The ability to track assets such as products, equipment, tools, and inventory in real-time is an essential component of AI manufacturing optimization in I4.0 [35]. A natural first step towards AI implementation is adapting the existing asset tracking system, transforming it into a real-time location system, and preparing it for the Industrial Internet of Things. The addition of RTLS to the IIOT environment of smart manufacturing can provide not only accurate asset tracking but also additional benefits for parts inventory, tool management, and personnel supervision [35].

With dataflow and product monitoring in place, various AI methods can be implemented to boost productivity, storage optimization, and sales.

Augmented reality has further been proposed as a disruptive and enabling technology within the I4.0 manufacturing paradigm. The term augmented reality (AR) has been defined as a system that “supplements the real world with virtual (computer-generated) objects that appear to coexist in the same space as the real world” [36]. Augmented reality may be used for troubleshooting and support. It allows experts to remotely assist in cases where they traditionally needed to be on location.

In contrast with maintaining equipment when the need arises, predictive maintenance, together with intelligent sensors, has also emerged as a new approach to maintenance in I4.0. Intelligent sensors make it possible to obtain an ever-increasing amount of data, which must be analyzed efficiently and effectively to support increasingly complex systems’ decision-making and management [37]. Using AI methods to predict when production machinery needs maintenance and for which parts, many benefits arise. For instance, predicting what spare parts need to be in stock at what time [37].

5.5 AI Potential in the Human Health Sector

The rapid innovation of analytics and data-driven technology within the human health sector is taking place worldwide. This holds significant potential for medical diagnosis by AI, natural language processing for assessing mental health [38, 39], and identifying high-risk and high-cost patients. Big data spaces offer the ability to establish an observational evidence base for clinical questions that couldn’t otherwise be answered. This could be particularly useful with issues of generalizability, allowing data-driven clinical decision support tools that may lead to cost savings and promote appropriate standardization of care. Clinicians could receive messages that inform them of the diagnostic and therapeutic choices made by respected colleagues facing similar patient profiles [40].

6 Discussion and Outlook

According to both our and the BN survey, many companies still find it challenging to incorporate AI, machine learning, and robotics, despite some progress. A slightly lower percentage of our respondents (9 of 21, 43%) compared to the BN survey (45.6%) answered that AI is complicated and it is hard to know where to start. Lack of understanding how AI can help their business may come from a lack of knowledge, but it can also be that AI can’t improve their business. Whereas 1 of 2 respondents in the SFJ agrees that they have plans for more data utilization, 2 of 3 respondents in the BN survey agree with the same statement. Interestingly, the BN survey found that leaders have a significantly lower understanding than employees regarding the criticality of AI and data for future business operations. However, our survey of businesses in Sogn og Fjordane showed the opposite result. Leaders have a better understanding of the criticality of AI in their company’s future business development compared to employees. Is it easier to convey change either in a bottom-up or top-down way in the district? Nevertheless, further research is necessary to determine if this difference exists for the region of Sogn og Fjordane, if it is a sample-size issue, or if in only one year the opinions of leaders have changed due to rapid changes in the markets and new understanding.

Although 12% less of the SFJ respondents compared to the BN survey agree that they are worried they don’t have enough competence to meet future technological challenges, a large number of companies in both surveys do not have regular competence training or finance for digital development. These aspects may be hindering enterprises from fully utilizing data-driven technologies and taking full advantage of their potential data. The absence of regular training to improve digital competency may also hinder the company’s ability to adapt and take advantage of new digital opportunities.

The study of various regions and their digital maturity is important since regions may have varied bases for the advancement of technology, adoption of digitization, and implementation of AI. The potentially large difference between the degree of leaders versus employees that understand how AI can change their business compared to the BN survey and our survey may indicate a regional difference. In smaller, more rural areas, the competition among businesses is generally less than in cities, and the work structure may be flatter due to smaller offices and less work-private distance. The latter may make it easier to convey change in a bottom-up or top-down way. The region also has some large and critical national digital services that are developed and operated by public entities located in Sogn og Fjordane (i.e., the digitalization directory in Norway). SMEs may be facing these challenges due to low capacity and a lack of financial and skilled human resources, preventing them from competing internationally and investing in training to ensure that employees are equipped with the skills and understanding required to increase the business’s competitiveness and robustness for future challenges.

Currently, there is limited research being conducted in companies located in Sogn og Fjordane, as evidenced by the low number of companies receiving support from Norwegian and European research funds. However, this does not necessarily indicate a lack of development in companies. Western Norway, which includes Sogn og Fjordane, is ranked highest in Norway for business-related development. However, development can also be achieved through investments in new machines, systems, and consulting services that have been launched in the market. Instead, networking, consulting with non-profit organizations, and investing in more research for innovation, growth, and development may be a good strategy to develop affordable skills internally, make collaborations across nations, and implement tailored AI solutions. Zooming out to a Nordic perspective on AI adaptation, key challenges found involve data management, transitioning from pilots to production, understanding regulations, and developing ethical and fair AI solutions, which are also likely aspects of regional challenges [12].

A generally low financial allocation for digital development may also indicate that companies may be reluctant to invest in these technologies and their training. Challenges faced among the surveyed companies may have more economical than “limited understanding of AI’s potential or digital readiness” aspects. Documentation of the impact of specific AI-use cases across sectors will likely be valuable towards a more predicable understanding of how data collection and AI may change businesses. Furthermore, incremental steps and a focus on cost-efficient actions will help gain knowledge and insights into data management and applicable AI by, for example, utilizing non-profit agencies for support and research engagements and collaborations. Clustering and networking with companies that face similar challenges may also be good strategies to bolster SMEs positions and effectively navigate the challenges that lie ahead. Companies that may utilize the same technology and that are not in direct competition with each other can benefit from co-joining projects and collaborating to learn from each other’s successes and failures. The company’s willingness to invest in AI technologies essentially depends on the economic output, which answers from this survey may also suggest: the company must generally be able to see a concrete positive impact and how big that impact is [41]. Such a realistic perspective is healthy, but there will also be cases where AI is not having enough impact to make a difference for the company or may in some way have a negative impact. However, few companies measure the success of AI implementation [13], and there is a great need for future research and companies to document the effects of AI adoption. Making more user stories and how-to’s with successes and failures of digital developments and AI implementation available in “safe/non-competitive” spaces can be valuable for faster and leaner change among SMEs.

7 Conclusions

The respondents see challenges in investing in AI but have a slightly greater understanding and ambition for it. We found that most companies surveyed have concrete plans to tackle future technological challenges, but not necessarily for data utilization. Our survey may indicate that enterprises in SFJ are not performing less regarding digital maturity, capacity, and human resources than enterprises in the larger western Norway more urban regions. Our survey, in fact, shows a better understanding among the managers and leaders surveyed. “The more you know, the more you don’t know”, is a common expression. Seeing potential and seeing what one must mobilize to solve a challenge is maybe a better way than thinking you have everything you need. The findings of the survey seem generally to be in line with other large surveys in the western Norway region, the Nordics, and Germany, which is positive news for a rural but rich region.

The two quantitative surveys in Vestland need to be followed up on with a new BN survey that better incorporates SFJ and rural areas as well as with a number of qualitative interviews that focus on the topics in-depth. The information provided by this strategy will be more accurate and thorough, and stakeholders can use it decision-making processes.