4. Digital innovation: Cross-sectoral dynamics and policy implications

Dominique Guellec, Caroline Paunov and Sandra Planes-Satorra

Directorate for Science, Technology and Innovation, OECD

With a focus on the agri-food, automotive and transportation, and retail sectors, Chapter 4 explores the impacts of digital transformation on innovation and identifies sector-specific dynamics. In view of such impacts, the chapter evaluates how innovation policies should adapt to promote vibrant and inclusive innovation ecosystems effectively. Examples of novel innovation policy approaches implemented in various countries are provided. The chapter also synthesises key findings from the OECD’s Working Party on Innovation and Technology Policy and, specifically its Digital and Open Innovation project. It explores in detail the changes needed in innovation policy in the digital age, considering the impacts of the digital transformation on innovation across sectors.

    

Introduction

Digital transformation is a multifaceted phenomenon that is impacting innovation in all sectors of the economy. New digital technologies, including artificial intelligence (AI), have enabled the creation of completely new digital products and services and the enhancement of traditional ones with digital features. Production processes are also subject to substantial change, with new modes of human-to-machine interaction (see Chapter 5). New opportunities are emerging across innovation processes – from research (e.g. the use of big data analytics, large-scale computerised experiments), to development (e.g. new techniques of simulation and prototyping) and commercialisation (e.g. use of marketplace platforms).

Despite the dramatic changes, the impacts of digital transformation on innovation in specific sectors are largely unknown or anecdotal. Since industries significantly differ in their products and processes, their structures and in how they innovate, the impacts of digitalisation on innovation are also likely to differ. For instance, end products in primary sectors such as food or mining remain largely unchanged. Conversely, the media, music and gaming industries, to name a few, have completely digitised their product and service offering. Another example is the wide deployment of robots in the automotive industry, while automation remains at early stages in sectors such as agriculture and retail. There is, however, little systematic evidence about the sector-specific impacts of digital transformation on innovation. Understanding these differences matters for policy aimed at supporting innovation systems, because countries’ industry composition differs markedly.

How is the digital transformation changing the innovation practices of firms?

This section explores how digital transformation is changing the nature of business innovation. Digital technologies have lowered information-related production costs and increased the “fluidity” of innovative products. Digitised knowledge (i.e. knowledge that takes the form of data) and information can circulate and be reproduced, shared or manipulated instantaneously by any number of actors regardless of their location. As a consequence of changes in costs and fluidity, four trends affect innovation practices across all sectors of the economy in the digital age (Figure 4.1).

Data are a key input for innovation

Data are increasingly used in innovation processes. They help explore new areas of product and service development. They help gain critical insights about market trends, consumer demand and the behaviour of competitors. And they optimise development, production and distribution processes; tailor the product and service offering to specific demands; and rapidly adjust to changes in demand. The emergence of smart and connected products, as a result of increased sensing, connectivity and data embedded in products, significantly contributes to the generation of new data.

Data have allowed the development of completely new services and business models. These have been enabled by the availability, and capacity to exploit, large amounts of real-time data. Examples include smart farming services, peer-to-peer accommodation services (e.g. Airbnb), on-demand mobility services (e.g. Uber), and platforms to search, compare and book accommodation and transportation options (e.g. Booking).

Business data are increasingly used to optimise processes within firms, but also within supply chains. Manufacturing sectors exploit abundant real-time shop-floor data to identify patterns and relationships among discrete processes. This allows manufacturers to optimise data by reducing waste, saving energy, increasing flexibility and better using assets, among other areas (OECD, 2017). For example, UPS, a multinational logistics company, uses a fleet management system enhanced by data analytics. It allows for route optimisation, increasing the efficiency and flexibility of delivery processes and reducing fuel consumption. Data are also used to predict the needs of production systems, significantly lowering maintenance costs compared to unplanned maintenance and repair. In agriculture, data from a multiplicity of sensors can help farmers optimise use of water and other inputs to boost yields.

Figure 4.1. Characteristics of innovation in the digital age

Source: OECD (2019), Digital Innovation: Seizing Policy Opportunities, https://doi.org/10.1787/a298dc87-en.

Services innovation enabled by digital technologies

Digital technologies offer opportunities for the creation of entirely new digitally enabled services. Predictive maintenance services, for example, use the Internet of Things (IoT), which involves the deployment of sensors and actuators connected to software systems. Other emerging services include on-demand transportation (e.g. Uber); and web-based businesses. New digital technologies have also propelled expansion of the sharing economy and greater customisation. Renting-as-a-service models can replace selling of equipment, for example, while businesses can harness software and data to adapt products to customers’ specific needs.

Such changes also contribute to a blurring of boundaries between manufacturing and services innovation. On the one hand, manufacturing firms increasingly offer innovative services to complement goods – a process known as “servitisation” of manufacturing. For instance, John Deere, an agriculture machinery producer, has developed a software platform that provides farm-management support services based on sensor data. On the other hand, service providers increasingly invest in digital technology to improve their activities. For instance, big retailers invest intensively in data collection and analytics capabilities (e.g. to personalise promotions and predict consumer trends), augmented and virtual reality (VR) (e.g. to develop digital fitting rooms) and the IoT (e.g. to improve inventory management).

Innovation cycles are accelerating

Digital innovations (such as three-dimensional [3D] printing and increasingly sophisticated simulations) introduce new and rapid innovation cycles by, among other routes, accelerating the processes of product design, prototyping and testing. Engineers and designers across manufacturing industries increasingly use “digital twins” (i.e. a 3D VR version of a production process or a product) to experiment with designs.

New technologies also stimulate market launch of testing (beta) versions that are regularly updated to incorporate consumer feedback. This is common practice for software launches. Many firms are also adopting a “lean start-up” method, which consists of creating minimum viable products that can be brought to market. Once launched, producers collect feedback from users and integrate it into the next development round. For example, GE Appliances’ FastWorks system, based on lean innovation principles, involves consumers early in the development of new products such as refrigerators (General Electric, 2017).

Innovation is becoming more collaborative

Innovation ecosystems are becoming more open and diverse. Firms interact with research institutions and firms due to three reasons. First, they gain access and exposure to a richer pool of expertise and skills complementary to their own competences (e.g. data analytics). Second, such collaborations allow sharing of the costs and risks of uncertain investments in digital innovation. Third, reduced costs of communication allow greater interaction among actors engaged in innovation (e.g. firms, public research institutions [PRIs]), regardless of their location.

Collaborations take different forms, including the following:

• Data sharing. The non-rivalrous nature of data allows various actors from different organisations to use the same database simultaneously, even if they are located around the world. This has stimulated firms to share their data for research and innovation purposes, often with universities and research organisations, or trusted business partners. Challenges and hackathons are other popular tools for sourcing external ideas to foster data-driven innovation.

• Partnerships. Partnerships with large technology firms, digital start-ups and PRIs are becoming more common in the digital age. Their goal is to join efforts to foster joint value creation, expand market potential and combine strengths. In so doing, they allow the closing of skills or competence gaps. Collaborations with digital start-ups, in particular, have also boomed in recent years. Such collaborations are seen as “digital accelerators”, with the flexibility needed to develop new disruptive technologies (Lund, Manyika and Robinson, 2016).

• Platforms. Industry platforms are products, services or technologies created by one or several firms. They provide the foundation upon which different actors can innovate by developing complementary products, services or technologies using digital tools (Gawer and Cusumano, 2014). These platforms can thus serve as the effective industry standard. They make development processes more efficient and less costly, and reduce time-to-market for new products. An example is the SmartDeviceLink Consortium, an open-source platform for smartphone app development for vehicles, created by Ford and Toyota. Firms also use crowdsourcing platforms to source ideas from outside the organisation (either the general public or a pool of accredited experts). In so doing, they aim to solve a specific problem or challenge, including finding new product or design ideas. Such initiatives are often conducted through intermediary platforms, such as InnoCentive.

• Acquisitions. The acquisition of innovative firms (particularly start-ups) by established firms is also a channel for collective innovation. Start-ups play a role in discovering and testing new markets and business models. When successful, they can be acquired by larger firms with access to capital and marketing channels that can help to scale a successful product.

The impacts of the digital transformation on innovation across sectors

How are digital technologies integrating different sectors?

Digital technologies are integrating and transforming sectors in different ways. This section explores transformations in the agriculture, automotive and retail sectors, which are considered representative of primary, secondary and tertiary sectors more generally.

Agri-food sector

In agriculture, intelligent and digitally connected machinery (IoT) enables the development of “precision farming”. This allows systems that help farmers improve the accuracy of operations. Such systems can also optimise the use of inputs (e.g. water, fertilisers, pesticides) to give each plant (or animal) exactly what it needs to grow optimally. Tractors and other agricultural machinery are equipped with a large number of sensors that capture information related to crops (e.g. soil conditions, irrigation, air quality, presence of pests). Drones equipped with sensors are also increasingly used for crop scouting and spraying. Data captured by in situ sensors, drones and satellites allows better monitoring of crop health, assessment of soil quality and optimisation of input use, thus having positive effects on productivity.

The introduction of robots is another trend in farming. Fruit-picking, harvesting and milking are examples of the repetitive and standardised tasks performed by agricultural robots. Robots also generate data that can be exploited for different purposes. For instance, Lely Industry, a manufacturer of milking robots, collects data from robots to exploit information regarding the feed, animal health and milk quality of individual cows (Lely, 2016). Although agri-robots are generally in early stages of development, they are expected to increase efficiency and allow for more automated and precise agricultural practices.

Large agriculture machinery producers and input suppliers, such as John Deere, are using large amounts of data collected through the IoT from farm applications and robots. They combine them with other data such as on the weather or markets to develop “smart farming” services. These use big data analytics and AI to inform farm-management decision making (Wolfert et al., 2017). Such systems can help farmers decide when to plant or harvest, to choose the type of crop to plant depending on soil conditions and market prices, and to automatically instruct agricultural robots to perform certain tasks. Precision and smart farming are still mainly restricted to large producers. Smaller producers are less likely to adopt precision farming technologies due to the costs of investment, and of learning how to use them and adapt production processes.

The agri-food supply chain is starting to use the IoT to trace the origins and track the whereabouts of products, as well as their transportation and storage conditions. In this way, it improves transparency in the value chain. Blockchain and other distributed ledger technologies are also expected to offer opportunities for increasing the traceability of food products from harvest to point of sale. Major food companies are collaborating with IBM to apply blockchain to make food supply chains more transparent and traceable and to streamline payments (Tripoli and Schmidhuber, 2018).

Automotive industry

In the automotive sector, rapid developments in digital technology are completely reshaping the industry. These include vehicle innovations (e.g. car connectivity, autonomous driving), innovations in production (with smart factories or Industry 4.0 applications) and new business models (with the provision of after-sales services and expansion into on-demand mobility services).

Digital technology has given rise to connected cars that generate data from the physical world, receive and process data, and connect to other cars and devices. Connected cars allow for enhanced driver safety and convenience. New services include automatic emergency calls after an accident and real-time road hazard warnings for drivers, car repair diagnostics and systems of time-saving networked parking. In addition, navigation systems optimise route planning by considering real-time traffic conditions.

Developments in autonomous driving are being propelled by advances in the fields of robotics, AI, machine learning (ML) and connectivity. There are five different levels of automation – from driver assistance to complete automation. All new car models offer driving assistance systems. These take over parts of the vehicle motion control and support the driver with certain tasks such as parking and speed-keeping – but the driver is still in charge of driving. From a technical viewpoint, technology for highly automated driving in controlled environments is quite mature (VDA, 2015). At full driving automation, cars drive independently and react to their environment without intervention of the driver. Such systems are being tested in pilot projects (PSC/CAR, 2017), but opinions differ greatly on when full autonomy might be achieved.

The automotive industry is also a leader in developing “smart factories”. It is adopting a variety of Industry 4.0 applications, including Internet-connected robotics, data analytics, and cloud and high-performance computing (HPC), among others. For instance, Hirotec, a Japanese auto parts manufacturer, uses ML and data analytics to predict and prevent failures. This drastically reduces the cost of unplanned downtime (Hewlett-Packard, 2017). BMW has set the goal of knowing the real-time status of all important machines producing components from all their suppliers using IoT applications (Ezell, 2018). Kern and Wolff (2019) provide other examples of investments by carmakers and automotive suppliers to foster efficiency, and automate production and supply-chain processes.

Firms in the automotive industry are also providing new services related to their products. Three areas of focus are the provision of new after-sales services (e.g. predictive maintenance); the development of alternatives to car ownership (e.g. vehicle subscription services); and expansion into on-demand mobility services (e.g. creation of own car-sharing brands).

The retail sector

In the field of retail, digital innovations aim at enhancing the consumer experience (both in physical and online shopping) and optimising processes (e.g. logistics, warehouse management). The largest investments focus on data collection (e.g. purchasing and browsing data) and data analytics capabilities. Such data provide insights on consumer needs and preferences that are used to customise the shopping experience, for instance by sending personalised advertisements and promotions. For example, Sephora uses data from customers’ online shopping histories by employing beacons in their stores. These beacons send smartphone notifications when customers near an item they had previously added in a digital shopping cart (Pandolph, 2017).

Innovations in physical stores are expanding. Smart dressing rooms, for instance, might recommend specific items of clothing. Digital mirrors can allow customers to try on and compare several outfits, among other things. And automatic payment systems allow customers to skip check-out lines. AmazonGo, for example, recently established a cashier-free store in Seattle. By deploying sensors, cameras and other digital technologies, the store allows for automatic payment of products that customers take off the shelf, without the need to scan bar codes (Amazon, n.d.). Innovations in online retail include applications for designing or personalising products (e.g. shoes) through 3D visualisations. The automatic reordering of products may also become more common. The Amazon Dash Replenishment Service, for example, allows connected devices (e.g. washing machines, coffee machines) to reorder products automatically (e.g. laundry detergent, coffee beans) when supplies are running low. However, all of these innovations remain marginal and are mainly deployed by large retailers.

The retail sector is using the IoT and robotics to better manage inventories (e.g. in warehouses) and optimise other processes. AI is also opening avenues for predictive analytics to strengthen forecasting and improve stock management. For example, Otto, a German online retailer, uses consumer data and a deep learning algorithm to predict what customers will buy a week before they order. The algorithm, which has 90% accuracy, has led Otto to introduce an innovative stock management system that automatically purchases products from third-party brands (The Economist/Capgemini, n.d.).

Why are the implications of digital transformation likely to differ across sectors?

The implications of the digital transformation likely differ across (and within) sectors due to a range of factors that can be grouped along three dimensions described below.

Opportunities for innovation using digital technology

Future technology developments are inherently uncertain. Yet given the important variety in the nature of a sector’s products and processes, some sectors will likely be disrupted to different extents by specific digital technologies (e.g. AI, IoT, drones, VR, 3D printing). Similarly, the transformation will probably take different forms and develop at different speeds. Depending on sectoral characteristics, digital technologies offer different opportunities for the following:

• Digitalising final products and services. While some industries have completely digitised their products over past decades (e.g. the media, music and gaming industries), others remain mainly physical, such as food and consumer products. Many industries present a mix of digital and physical components in their final products, with the digital ones often becoming progressively more important. In the automotive industry, vehicles increasingly integrate digital features. Advanced infotainment systems and other functionalities enabled by connectivity and data analytics, for example, are becoming key considerations in consumer purchasing decisions.

• Digitalising business processes. The extent to which digitalisation affects sectors’ business processes may differ. It depends on the nature of the activities and the characteristics of production (e.g. whether it involves the assembly of physical products, if the sector is characterised by long supply chains, etc.). In particular, digital technologies offer opportunities for digitalisation (and automation) of production processes; for interconnecting supply chains; and, for improving interactions with the final consumer.

• Creating new digitally enabled markets and business models. New markets or market segments enabled by digital technologies, often adjacent to traditional sectors, have been created over recent years. E-commerce, car-sharing services and Fintech services are well-known examples. While new business models are emerging across the economy, the scale and disruption potential of these trends vary across sectors. In some cases, those business models may displace traditional ones (e.g. travel agencies). In other cases, the two models may co-exist and expand the product or service offering (e.g. brick-and-mortar existing simultaneously with online retail stores).

Data needs and challenges for innovation

Data have become a key input for innovation (Table 4.1). However, sectoral differences also arise because access to data differs across sectors. For instance, in some sectors, data needed for innovation are more sensitive than in others (such as patient data for healthcare innovations). They may also be less widely available (such as farming data for the development of smart farming services, given the low digital technology uptake in agriculture). The nature of data privacy and safety challenges also differ. The protection of data generated by connected cars and transportation systems is critical to avoid cyber-attacks that could put road safety at risk. At the same time, misuse and leakage of personal data are more problematic in the retail sector. Some sectors may also be more attractive than others to digital talent, creating differences in the capacity to exploit data. Data ownership conditions may likewise be a barrier to innovation in some sectors. This is particularly the case in sectors like agriculture where data are captured by some actors but exploited by others.

Unequal access to data across firms can create an uneven playing field within the same sector. This, in turn, can contribute to higher market concentration. Amazon and Google, for example, have higher capacity to access large amounts of consumer data compared to other retailers.

Table 4.1 presents some of the differences across the agri-food, automotive and retail sectors regarding the type of data needed for innovation purposes, and the opportunities and challenges related to those data types.

Table 4.1. Differences in data requirements for sectoral applications
	Data needs	Main challenges
Agriculture (precision agriculture)	● Aggregated sensor data from many farms (captured by sensors in the fields or mounted on machinery or drones) ● Satellite data (GIS, meteorological data, satellite imagery on crops)	● Low levels of digital technology adoption and high cost of uptake, particularly for small farms ● Data sharing (resistance by farmers) ● Data quality and integration ● Building trusted data analytics
Automotive industry (connected cars)	● Data on driver behaviour, car status and location ● Historical data on car performance (for predictive maintenance services) ● GIS, real-time traffic information	● Skills to exploit data ● Data integration ● Data privacy (risks, e.g. usage-based insurance contracts) ● Road safety (risk of cyber-attacks)
Retail (personalisation of consumer experience)	● Customers and transactions data ● Personal data on social media and search websites	● Skills to exploit data ● Data integration ● Personal data privacy (risk e.g. price discrimination)
Note: GIS = geographic information system.
Source: Paunov and Planes-Satorra (2019), “How are digital technologies changing innovation? Evidence from the agriculture, automotive and retail sectors”, https://doi.org/10.1787/67bbcafe-en.

Digital technology adoption and diffusion trends

Digital technology adoption is heterogeneous across sectors (Calvino et al., 2018). Industry estimates, for instance, show that sectors such as automotive and financial services are leading AI adoption, relative to the tourism and construction sectors, among others (Bughin et al., 2017). Key factors influencing adoption include:

• Capabilities to take up new digital technologies. Skills for adoption differ across sectors. For instance, sectors such as agriculture and construction, characterised by relatively high shares of low-skilled workers, register low uptake for digital technology. Capacities needed for digital technology adoption include skills at the individual level (e.g. information and communication technology skills, data expertise or previous related knowledge) and at the organisational level. The latter include the capacity to fine-tune organisational structures, adjust processes, redefine strategies and tasks, and manage emerging risks, among others. In this sense, the capacities of managers and their understanding of digital transformation dynamics are critical.

• Pressures from market competition. The emergence of new players in the market (i.e. digital start-ups or tech firms that enter existing markets or create new activities adjacent to traditional sectors) is pushing incumbents to innovate. However, such pressures seem to be more critical in some sectors than others. For instance, in the automotive industry, the market entry of firms such as Alphabet (investing in the development of self-driving cars) and Zipcar (offering car-sharing services) is pressuring incumbents to embrace new digital innovations.

  Some sectors are particularly affected by the emergence of new platforms. For example, the emergence of digital platforms is significantly reshaping the tourism industry. Booking.com, for example, enables consumers to search, compare and book accommodation and transportation options. As another example, sharing economy platforms such as Airbnb provide for peer-to-peer accommodation services.

• Specific sectoral characteristics and structures. Sectoral characteristics also influence the pace of digital technology adoption. Digital technologies, in particular, permeate the activities of different types of actors within the sector. These range from small and medium-sized enterprises (SMEs) to large firms, start-ups and research institutions. Large firms are usually early adopters of new technologies. This is mainly due to their access to the resources needed to invest in new technologies and the greater presence of workers with relevant technical expertise. However, large firms may also suffer from inertias, hierarchical and rigid structures, and legacy systems that can hamper their transformation (Rogers, 2003; Zhu, Kraemer and Xu, 2006).

  Firms integrated into global value chains may be more exposed to digital technologies, and have higher incentives to adopt them. Their suppliers may adjust more rapidly to requests from upstream producers to adopt new practices, and receive support to implement them. For instance, Toyota supports its suppliers in implementing their new production systems (Kern and Wolff, 2019). The diffusion of digital technology also relies on access to critical infrastructure, such as broadband Internet connection. This may be a challenge for sectors and firms located in more remote or rural areas. Firms in agriculture are a case in point.

• Changes in consumer demands. Changes in consumer needs and demand are also driving the digital transformation of sectors. For example, in the field of transportation, younger generations (especially in urban areas) show a higher preference for on-demand schemes, rather than car ownership. In retail, consumers show increasing preference for a combination of physical and online shopping, along with the quick delivery of products purchased on line.

• Level of resistance to change. Resistance to change may also differ across sectors, depending on several factors. First, resistance may correlate to awareness of the opportunities offered by digital technologies. It could also depend on the perceived and actual challenges for specific stakeholders from adopting digital technologies, including job displacement or retraining requirements. Finally, it could depend on absorption capacities and the state of development of sector-specific digital technology applications. Low levels of technology adoption may also reflect consumers’ resistance to change, which differs across products. For instance, the adoption of e-commerce was initially slow, while user rates of mobile transportation services differed strongly across countries. Similarly, there may be more resistance to accepting robots for personal care services than for new transportation services.

How should innovation policy be adapted to the digital age?

Effective policy for innovation in the digital age requires governments to adopt policy mixes that respond to the changing context created by the digital transformation. The new mix should comprise five key policy objectives (Table 4.2): ensuring access to data for innovation; providing support and incentives to innovation and entrepreneurship; building a strong public research system and having a skilled labour force; fostering collaborative, competitive and inclusive innovation ecosystems; and setting national policies that account for the global context and citizens’ concerns.

Digital transformation calls for changes that affect all innovation policies, but to varying degrees. Some domains of policy need to adapt their target or content to digital innovation, while essentially preserving their core objectives. This includes, for instance, policies supporting entrepreneurship, digital technology adoption by SMEs and the development of general-purpose technologies. Other domains need more change, including rethinking of the policy rationale: that includes public research policy (moving towards open science).

Access to data has become a major new theme in all policy domains relating to innovation, such as support for business innovation, public research and competition policy. Data has also become a policy domain itself, and subject to issues such as confidentiality and privacy that directly impact innovation.

Policy makers need to address several new challenges. These include ensuring greater responsiveness and agility of policies and setting national policies in view of developments in global markets. They must also provide information and foster dialogue so that citizens are well-informed of the realities of new technologies and can participate in choices over funding of technologies considered harmful. Finally, they must ensure that government can access the advanced skills (e.g. in the field of AI) and data needed to design appropriate regulations and policies, ensuring that new technologies do not harm the public interest.

A sectoral approach is required in policy areas in which sectors have different challenges and needs. This is particularly the case of data access policies, digital technology adoption and diffusion policies, and support for digital technology application development. For example, challenges in agriculture often relate to data sharing and integration, while in retail ensuring data privacy is a rising concern.

Table 4.2. Major changes to innovation policies called for by digitalisation, by policy domain
Policy domain	Changes required
Access to data	● Ensure access to data for innovators, taking into account the diversity of data and preserving rights and incentives1 ● Explore the development of markets for data
Business innovation and entrepreneurship	● Ensure that policies are anticipatory, responsive and agile ● Support service innovation that implements digital technologies ● Adapt the intellectual property (IP) system ● Support the development of generic (multi-purpose) digital technologies1
Public research, education and training	● Promote open science (access to data, publications) ● Support training in digital skills for science ● Support interdisciplinarity in research ● Invest in digital infrastructure for science ● Facilitate co-creation between industry, science and civil society ● Ensure that skills needed for digital innovation are being developed (in collaboration with education and labour market policy authorities) ● Support education and training for the development of managerial skills
Competition, collaboration and inclusiveness	● Review competition policies from the perspective of innovation in the digital age (e.g. rules regarding takeovers and standards; IP systems) ● Support digital technology adoption by all firms, particularly SMEs1 ● Support social and territorial inclusiveness in digital innovation activities
Cross-cutting principles	● Frame national policies in view of global markets ● Engage with citizens to address technology-related public concerns ● Adopt a sectoral approach to policy making when necessary
← 1. These areas require a sectoral approach to innovation policy making.
Note: SMEs = small and medium-sized companies.
Source: OECD (2019), Digital Innovation: Seizing Policy Opportunities, https://doi.org/10.1787/a298dc87-en.

Innovation is also influenced by many policies that do not target innovation explicitly or primarily. These include education, tax, health, environmental, transportation and competition policies. Competition policy is particularly critical for innovation, as only the right competitive environment will stimulate firms to innovate and foster innovation-driven growth.

Data access policies

As data now constitute a major input to innovation, access to data – and to the tools that gather and help interpret data – will influence who participates in digital innovation, and in what ways. Therefore, a specific policy agenda around data access needs to be developed (OECD, 2015a). The main objective of data access policies should balance two elements. On the one hand, policies should ensure the broadest possible access to data and knowledge (incentivising sharing and reuse) to favour competition and innovation. On the other, they should respect constraints regarding data privacy, ethical considerations, economic costs and benefits (i.e. incentives to produce the data) and intellectual property rights (IPRs).

Policies should consider the diversity of data types, which imply differences in terms of access and other challenges associated with their generation, access and exploitation. Access to public research data, in particular, allows the reproduction and testing of the validity of scientific research, as well as reuse in further research (OECD, 2015b; Dai, Shin and Smith, 2018). Some governments establish open access to data generated by public services (e.g. weather monitoring, urban transportation, etc.) to foster data-driven innovation. For instance, the United Kingdom’s open data portal (data.gov.uk) publishes data from the central government, local authorities and other public bodies. It produces data on a variety of fields to create new opportunities for organisations to build innovative digital goods and services. Other examples include the open data portals of Canada (open.canada.ca), France (data.gouv.fr), Japan (data.go.jp) and the United States (data.gov).

Appropriate conditions should also be created to allow for the emergence of data markets. Trading data may facilitate innovation, as well as put a price on data generation and curation for future use – thus facilitating the generation of more data. Markets in data also facilitate entry for start-ups that are data-poor but which require data as part of their business model.

There are, however, major challenges to the development of markets for data (and knowledge). Data, for example, are often adapted to a specific context. Outside of this context, they may have little or no value, thereby limiting their transferability. Other key challenges relate to appropriability, difficulties in evaluating the true market value and quality of data, and privacy and safety concerns affecting personal data.

Policies to support innovation and entrepreneurship

Ensure that policies are anticipatory, responsive and agile

The policy instruments needed for the digital age should be anticipatory, responsive and agile. The innovation agenda is shifting quickly and difficult to predict in certain fields. Therefore, government needs to become more flexible and alert to change, while keeping (prudential) rules of engagement when it comes to specific policy instruments.

Approaches to ensuring this policy responsiveness includes the deployment and monitoring of small-scale policy experiments. These experiments can help assess their relevance and efficiency in a context of high uncertainty, based on which they could be easily scaled up, down or abandoned.

In a context of rapid change, it is also critical to streamline application procedures for innovation support instruments. For example, the Pass French Tech programme offers fast-growing start-ups simplified and quick access to services to help them expand. These include services in areas such as financing, access to new markets, innovation and business development (French Tech, n.d.).

Using digital tools to design innovation policy and monitor policy targets is another option to spur faster and more effective decision making. For example, semantic analysis can identify policy trends and anticipate technology trends by exploring large quantities of textual information (e.g. innovation policy documents, patents, scientific articles) (OECD, 2018). While still experimental, semantic analysis has tracked strengths in specific research fields based on text information contained in publications. It is also used by innovation and research funding agencies to build better connections between recipients of support, based on information on their research activities.

Instruments that do not target a specific technology can also increase flexibility. Mission-oriented programmes that set a goal, but do not impose the means to reach it, can help. Such programmes may provide more autonomy and agility to choose the proper technological avenues to achieve a stated policy objective. The drawbacks of instruments without a specific target must be considered against the advantage of greater flexibility.

Certain environments, including public procurement with specific requirements such as data security, leave no choice of technology. In these cases, designing public institutions connected to technology developments in the private sector can prove useful. These institutions inform governments about the latest technology developments, as well as their potential benefits and harmful impacts. Data61 in Australia and the Digital Catapult in the United Kingdom are examples.

Support service innovation to lever the potential of digital technologies

Many innovation policies have been conceived for innovation in manufacturing. This has specific characteristics, such as being intensive in research and development (R&D), and often resulting in patents. In a context where services are becoming a key focus of innovation, policy initiatives should ensure that services innovation is considered. Initiatives that include emerging needs in services innovation may include support for projects. These could develop entirely new services using digital technologies, such as the Smart and Digital Services Initiative in Austria (FFG, n.d.) It could also support manufacturing SMEs to develop services related to their products. Service design vouchers for manufacturing SMEs in the Netherlands is one such example (RVO, 2018).

Adapt intellectual property systems

Digitalisation is transforming the IP system, which was designed for tangible inventions embodied in physical products and processes. With digitalisation, the IP system is confronted with new questions that require policy responses. These include how to incentivise data generation in a context of increasingly open data systems. Other questions revolve around ownership of patentable inventions produced by AI and the risk of counterfeiting the intangible components of products.

New digital technologies may also help enforce IP rights. Eventually, blockchain-enhanced IP on a range of intangible goods (e.g. photographs, music, movies), and even some tangible goods (such as 3D-printed items with unique digital identifiers), may help to create a more easily enforced system of IP.

Support development of generic digital technologies to respond to societal challenges

Policies need to ensure that multi-purpose digital technologies are developed to serve both commercial purposes, as well as social and environmental purposes. Public research is often driving advances in these areas, while an increasing number of non-profit organisations and private firms also pursue such objectives. Many examples exist of AI applications that tackle social and environmental challenges. Satellite imagery and deep learning techniques, for example, identify illegal fishing vessels and monitor changes in coral reefs to inform conservation interventions. Audio sensors can detect illegal logging. And face detection, social network analysis and natural language processing can identify victims of sexual exploitation on the Internet (Chui et al., 2018).

More engagement and debate with the public is also needed to demonstrate the characteristics of these technologies and appropriately address public concerns (e.g. privacy protection, development of applications for the public good). A lack of engagement with society creates the risk of a future backlash. This could have negative impacts on the development and deployment of important technologies.

Public research and education policies

Promote the digitalisation of public research

Strengthening researchers’ digital skills would ensure that new digital tools are integrated into public research processes (e.g. ML techniques). Specific training and capacity building activities, for example, could be offered. This tactic is a key objective of the digitalisation strategy for the higher education sector in Norway (2017-21) (Government of Norway, 2018).

Such measures should be accompanied by investments in digital tools and infrastructures critical for research (e.g. platforms for data sharing, supercomputing facilities for AI). Japan, for example, is investing more than USD 120 million annually to build a HPC infrastructure. It will be accessible to universities and public research centres for R&D purposes in a range of fields (HPCI, n.d.).

Stimulating interdisciplinary research (e.g. cross-departmental research projects) and the engagement in partnerships with other research institutions and with industry is also a larger priority in a context of digital innovation. Specifically, data science applications provide for new opportunities across academic disciplines. With regards to partnerships between industry and science, physical spaces remain important for more collaborative innovation. However, digital platforms can complement physical space and allow for new types of collaboration across geographic boundaries.

Build digital skills, including in the field of data analytics

Education and research authorities play a key role in building the digital skills needed across the economy. Innovation authorities should collaborate with them towards several goals. First, they help identify the new skills needed in a context of digital transformation. Second, they should provide inputs for university and vocational training programmes to fill critical talent shortages (e.g. data scientists) that often requires more interdisciplinary curricula. Innovation authorities also facilitate training for SMEs from traditional sectors to ensure they leverage the potential of digital technologies.

Foster competitive, collaborative and inclusive innovation ecosystems

Ensure that innovation ecosystems remain competitive

Dialogue between competition authorities and agencies responsible for innovation policy should address key questions. These include the use of data as a source of market power and the contestability of markets in which digital innovation is an important feature. Such markets are subject to rapid innovation (a source of contestability) and various sorts of scale economies (a source of persistent concentration) (Guellec and Paunov, 2018). New policies need to recognise the importance and prevalence of economies of scale, while ensuring equal access to markets and resources. As competition in digital markets is global, greater co-operation across jurisdictions may also be needed (OECD, 2019).

Do innovation policy instruments and regulations (e.g. support for R&D, IPRs) have an asymmetric impact on market players? Policy makers should consider this question. While such instruments are accessible to all in principle, this may not be the case in practice. For example, firms may lack capacity to defend their IP rights in courts, to co-operate effectively with public labs or to access public procurement.

Support collaboration for innovation

In the digital context, innovation policies will have to continue supporting collaborative innovative ecosystems. New policy approaches to foster collaborative innovation include the use of crowdsourcing and open challenges, as well as the creation of living labs. These approaches can help find innovative solutions to pressing challenges and foster co-creation between various actors.

Intermediary organisations, such as the Fraunhofer Institutes in Germany and the Catapult Centres in the United Kingdom, have become central players in innovation ecosystems. They provide services such as matching firms that need technology solutions with potential suppliers. New research and innovation centres, often public-private partnerships, have also been created. These centres provide spaces for multidisciplinary teams of public researchers and businesses to work together on specific technology challenges. They often stand out for their innovative organisational structures. Examples include Data61 in Australia and Smart Industry Fieldlabs in the Netherlands (Box 4.1).

Box 4.1. Approaches to fostering collaboration for innovation in the digital age

Intermediary organisations

Intermediary organisations connect different actors in innovation ecosystems (innovators, big firms, SMEs, investors, etc.) and facilitate their matching and collaboration for research and innovation. The Catapult Centres in the United Kingdom are a network of ten not-for-profit, independent physical centres that connect businesses with the United Kingdom’s research and academic communities. Each focuses on a strategic technology area in which the United Kingdom has great potential for growth. They offer a space with the facilities and expertise to enable businesses and researchers to collaboratively solve key problems and develop new products and services on a commercial scale. They also support firms’ access to foreign markets, create and retain high value jobs and attract inward investments from global technology businesses.

Collaborative research and innovation centres

Several countries have created (networks of) research centres. Multidisciplinary teams of public researchers and businesses work together at these centres on specific technology challenges. The centres provide spaces for collaboration and co-creation and often stand out for their innovative organisational structures.

CSIRO’s Data 61, Australia’s largest digital R&D centre, aims to put Australia at the forefront of data-driven innovation. To that end, it pursues new-to-the-world fundamental and applied research and works collaboratively with others in the innovation ecosystem, including universities, government and industry. To increase agility and attract digital talent, Data61 has adopted a “start-up culture” or “market pull” approach. Organisational structures are flatter (i.e. with less middle-management and higher autonomy of staff). Research leaders are encouraged to experiment with new ideas and take risks, while maintaining alignment with the strategic goals of the organisation. A “challenge model” also stimulates multidisciplinary teams to address large-scale social and business challenges.

Smart Industry Fieldlabs in the Netherlands are public-private partnerships to create physical or digital spaces for member companies and research institutions. Together, they develop, test and implement new smart industry technological solutions in various fields. These include automation, zero-defect manufacturing, flexible production, value creation based on big data, 3D printing and robotics. The 32 field labs, which have flat structures and follow a project-based approach, typically include users of such solutions, (potential) suppliers and knowledge institutes. They are active in collaborative research, concept validation, prototyping, testing and validation.

Crowdsourcing, open challenges and living labs

Various countries are harnessing the power of crowdsourcing, open challenges and living labs to drive innovation. The US government designed Citizenscience.gov to enhance use of crowdsourcing to engage the public in addressing social needs and accelerate innovation. The Social Challenges Innovation Platform (SocialChallenges.eu) encourages social innovators and entrepreneurs to post innovative solutions to social and environmental challenges that public authorities, private firms or non-governmental organisations aim to solve. Pit Stops, organised by the Digital Catapult (United Kingdom), encourage open innovation by bringing together large firms, SMEs, start-ups and academics to solve specific technology challenges. Disruptive technology start-ups and other actors able to solve such challenges are identified via online open calls.

Living labs are localised areas of experimentation within urban environments in which stakeholders collaboratively develop new technology-enabled solutions. For instance, Antwerp (Belgium) is developing a “City of Things” (IMEC, n.d.) through installation of a dense network of smart sensors and wireless gateways in buildings, streets and objects. Companies can use collected data to build innovative applications.

Source: OECD (2019), Digital Innovation: Seizing Policy Opportunities, https://doi.org/10.1787/a298dc87-en.

Support digital technology adoption by all firms, particularly SMEs

Firms (particularly SMEs) face important challenges to adapt to digital transformation. Such adaptation requires much more than simply purchasing new computers and software: it is about changing business processes, and often business models. This frequently implies new strategic capabilities, new skills, investments in new technologies and significant restructuring, all of which can carry risk. The failure of many SMEs that do not digitalise would mean the loss of much industry and market-specific know-how, which constitutes unique intangible capital. It is therefore in the public interest to support adaptation of SMEs selectively; less competitive firms would not be saved to avoid hampering the competitive process.

Innovative policy to foster diffusion focuses on helping test new digital technology applications by, for instance, creating test beds and regulatory sandboxes. Innovative initiatives also enhance early adoption of advanced digital technologies. To that end, they help innovators access state-of-the-art facilities and expertise (e.g. in the fields of AI or supercomputing). SMEs are revisiting traditional instruments to foster technology adoption such as awareness-raising campaigns, innovation vouchers, technical assistance and training. They are adapting these instruments to specific challenges of the digital age, and often use digital tools themselves (Box 4.2).

Box 4.2. Demonstration and testing of new digital technologies

Demonstration and testing facilities for SMEs

Some countries have established new facilities for demonstration and testing of digital technologies to increase adoption. For instance, the SME 4.0 Competence Centres in Germany offer SMEs access to demonstrations of Industry 4.0 technologies and sector-specific applications (e.g. 3D printing, sensors). These demonstration facilities are often located at universities and allow simulation of business and production processes in a similar to real-world environment.

The Industry Platform 4 FVG, established in the Italian region of Friuli Venezia Giulia, offers access to testing equipment, prototyping tools and demonstration labs. Several Austrian universities (TU Wien, TU Graz and Johannes Kepler University Linz) have also set up pilot factories, where SMEs have the chance to test new technologies and production processes without affecting production in their own facilities.

Experimenting with new technology applications

Countries are also exploring novel approaches to fostering testing of, and experimentation with, new digital technologies and applications in a near to real-world environment:

• Test beds provide environments where new technology developments can be tested in controlled but near to real-world conditions. Such testing environments are critical for research and innovation in certain areas, such as autonomous driving, and help accelerate adoption of new digital technologies. Finland is establishing a number of test beds for the open development of transport and mobility solutions, including automated driving, mobility-as-a-service and intelligent traffic infrastructures. In the United Kingdom, the National Health Service (NHS) introduced a Test Beds Programme in partnership with industry. This allows testing of innovations such as combinations of new digital devices such as sensors, monitors and wearables with data analysis. It also permits testing of new approaches to service delivery facilitated by digital technologies. Successful innovations are then made available to the NHS and care organisations around the country.

• Regulatory sandboxes provide a limited form of regulatory waiver or flexibility for firms to test new products or business models with reduced regulatory requirements. At the same time, they preserve some safeguards such as ensuring appropriate consumer protection. Sandboxes help identify and better respond to regulatory breaches and enhance regulatory flexibility. They are particularly relevant in highly regulated industries, such as financial services, transport, energy and health. The United Kingdom’s Financial Conduct Authority pioneered this approach with the launch of the Fintech regulatory sandbox, which encourages innovation in financial technology. The sandbox provides a controlled environment for businesses to test innovative products and services without incurring the regulatory consequences of pilot projects.

Source: OECD (2019), Digital Innovation: Seizing Policy Opportunities, https://doi.org/10.1787/a298dc87-en.

Support social and territorial inclusiveness

Innovation policies play a role in enhancing participation of disadvantaged individuals in digital innovation activities. Policy instruments address social inclusiveness challenges in the digital age in various ways. Some aim to build capacities through, for example, digital skills and entrepreneurship education. Others address discrimination and stereotypes through role models and mentoring programmes, among other approaches. Still others address barriers to entrepreneurship faced by disadvantaged groups. For example, they facilitate access to finance through microcredit or equity financing, provide tailored business development support and promote entrepreneurs’ insertion in business and research networks. Planes-Satorra and Paunov (2017) provides a wide range of examples of inclusive innovation policy.

Digital transformation also seems to favour further concentration of innovation activities in innovation hotspots (often urban areas). This calls for policies favouring territorial inclusiveness. “Excellence-based policies”, even if blind to location, tend to favour geographical concentration, since excellence is concentrated. These indirectly widen the gap between leading and lagging regions. Excellence-based policies should therefore be complemented by policies favouring geographical inclusiveness and diversity. They should focus on fostering innovation at the local/regional level, and building on specific regional strengths and comparative assets (e.g. the Smart Specialisation approach in the European Union).

Cross-cutting policy principles

Set national policies in view of developments in global markets

Digitalisation facilitates the circulation of knowledge, including across national borders, reducing government’s ability to restrict the benefits of policies to its own country. That raises a challenge for national policy makers. How can they ensure their own citizens (and taxpayers) benefit from national policies? Furthermore, how can they ensure that most of the benefits (e.g. income generated, productivity gained or jobs created) do not leak abroad? One related question concerns the sharing of benefits generated by the exploitation of national data (e.g. from the public health system) by foreign multinationals. Co-operative solutions are needed to share benefits arising from international flows of data and knowledge linked to national policies among countries. The OECD activity on base erosion and profit shifting is a step in this direction.1

Engage with citizens to address technology-related public concerns

The digital transformation has captured much attention in the press and with the public. In some cases, people fear leakages of personal data and the threat of robots taking jobs. Government and other actors must engage with stakeholders about these technologies and allay concerns through, for example, enhanced data privacy protection. Consultations with the public during the development of digital transformation strategies and other related policies can contribute. Without such public engagement, there is a risk of backlash in the future. This could have potentially negative impacts on the development and deployment of these technologies and their related benefits (OECD, 2015b; Winickoff, 2017; Dai, Shin and Smith, 2018).

Adopt a sectoral approach to policy making when necessary

Three policy domains require a sectoral approach when designing new initiatives, as the challenges and needs faced by sectors in these areas vary significantly:

• Data access policies should consider the diversity of data types needed for innovation in different sectors, given differences to access and other challenges associated with data generation, exploitation and ownership. For instance, precision agriculture draws mainly on sensor and satellite data. Conversely, the retail sector exploits consumer purchasing and social media data to personalise services. In agriculture, challenges often relate to data sharing and integration, while in retail ensuring data privacy is a rising concern.

• Digital technology adoption and diffusion policies should be tailored to the specific needs of the sector and/or type of actor (notably SMEs). These policies could involve awareness-raising, training and education, demonstration and testing of new technologies, and the operation of intermediary institutions. Diffusion is more challenging in some sectors than in others due to different production structures. For instance, many small firms in a sector may be geographically dispersed or a few larger ones may be geographically close. Other challenges include the landscape of intermediary institutions and/or the availability of digital capacities.

• Policies to develop sectoral applications of digital technologies should be supported where market conditions have inhibited the development of private sector-led solutions. This will ensure that such technologies provide benefits across the economy. The gap between future digital technology opportunities and current applications differs across sectors. This frustrates adoption of digital technologies for firms operating in certain sectors where applications do not yet exist (e.g. in the field of AI). Public research could support building more applications and help adoption across the economy where private business does not have the incentives to produce them.

Designing effective and tailored support to sectors operating in the digital context requires, as a first step, establishing mechanisms to strengthen policy intelligence. These may include roadmaps or sectoral plans for strategic sectors, in collaboration with industry stakeholders and social partners. Examples include the Sector Competitiveness Plans developed by six sector-specific Industry Growth Centres in Australia (Government of Australia, 2017).

Conclusion

Chapter 4 discusses the impacts of the digital transformation on innovation processes and outcomes. The chapter highlights general trends across the economy and factors behind sector-specific dynamics. In view of such impacts, it evaluates how policy support to innovation should adapt and in what directions, providing examples of novel approaches to innovation policy.

The chapter shows that four pervasive trends characterise innovation in the digital age. First, data are becoming key inputs for innovation. Second, innovation activities increasingly focus on the development of services enabled by digital technologies. Third, innovation cycles are accelerating. Virtual simulation, 3D printing and other digital technologies are providing opportunities for more experimentation and versioning in innovation processes. Fourth, innovation is becoming more collaborative, given the growing complexity and interdisciplinary needs of digital innovation.

Impacts of the digital transformation differ significantly across (and within) sectors. This reflects differences in the scope of opportunities for innovation in products, processes and business models that digital technologies offer. It reflects differences in the types of data needed for innovation and thus the challenges faced for their exploitation. And it reflects different conditions for digital technology adoption and diffusion.

The effective development of innovation in the digital age requires that governments adopt policy mixes that respond to the changing context created by the digital transformation. The changes called for by digitisation affect the entire innovation policy spectrum, but to varying degrees across policies. Access to data has become a major new theme in all policy domains relating to innovation, such as innovation support, public research and competition. It has also become a policy domain in itself, subject for instance to confidentiality and privacy issues that also directly impact innovation.

New challenges for policy making that need to be addressed include ensuring greater responsiveness and agility of policies; setting national policies in view of global markets; and engaging with citizens on new technologies. Equally, policy makers must ensure that government can access advanced skills, such as in the field of AI, and data needed to design appropriate regulations and policies. Finally, they must ensure that new technologies and applications do not harm the public interest. A sectoral approach is also needed in some policy areas.

This chapter is a first step in understanding the changing characteristics of innovation in the digital age. An important priority for policy research in this field involves gathering cross-country information on adoption rates of most advanced digital technologies at the firm level. Such data need to consider and capture ongoing technology trends. They would allow adoption trends to be measured across sectors, and across types of firms and locations. This, in turn, would help better identify the specific factors spurring and restraining digital innovation.

References

Amazon (n.d.), “Amazon Go”, webpage, www.amazon.com/b?node=16008589011 (accessed 16 April 2018).

Bughin, J. et al. (2017), “Artificial intelligence: The next digital frontier?” Discussion Paper, June, McKinsey Global Institute, https://www.mckinsey.com/~/media/McKinsey/Industries/Advanced%20Electronics/Our%20Insights/How%20artificial%20intelligence%20can%20deliver%20real%20value%20to%20companies/MGI-Artificial-Intelligence-Discussion-paper.ashx.

Calvino, F. et al. (2018), “A taxonomy of digital intensive sectors”, OECD Science, Technology and Industry Working Papers, No. 2018/14, OECD Publishing, Paris, https://doi.org/10.1787/f404736a-en.

Chui, M., et al. (2018), “Notes from the AI frontier: Applying AI for social good”, Discussion Paper, December, McKinsey Global Institute.

Dai, Q., E. Shin and C. Smith (2018), “Open and inclusive collaboration in science: A framework”, OECD Science, Technology and Industry Working Papers, No. 2018/07, OECD Publishing, Paris, https://dx.doi.org/10.1787/2dbff737-en.

Ezell, S. (2018), “Why manufacturing digitalization matters and how countries are supporting it”, Information Technology and Innovation Foundation, Washington, DC, www2.itif.org/2018-manufacturing-digitalization.pdf (accessed 21 January 2019).

FFG (n.d.), “Service innovations – Funding, guidelines”, webpage, https://www.ffg.at/en/programme/service-innovations (accessed 26 June 2019).

French Tech (n.d.), “Pass French Tech”, webpage, www.lafrenchtech.com/en-action/pass-french-tech (accessed 25 May 2018).

Gawer, A. and M. Cusumano (2014), “Industry platforms and ecosystem innovation”, Journal of Product Innovation Management, Vol. 31/3, Product Development and Management Association, St. Paul, United States, pp. 417-433, https://doi.org/10.1111/jpim.12105.

General Electric (2017), What is FastWorks?, GE Reports Canada, www.ge.com/reports/tag/fastworks/ (accessed 25 April 2018).

Government of Australia (2017), Industry Growth Centres Initiative: Sector Competitiveness Plans Overview, September, Department of Industry, Innovation and Science, www.industry.gov.au/sites/g/files/net3906/f/May%202018/document/pdf/industry_growth_centres_initiative_-_sector_competitiveness_plans_overview.pdf.

Government of Norway (2018), “Digitalisation Strategy for the Higher Education Sector 2017-2021”, webpage, www.regjeringen.no/en/dokumenter/digitalisation-strategy-for-the-higher-education-sector-2017-2021/id2571085/sec5 (accessed 29 May 2018).

Guellec, D. and C. Paunov (2018), “Innovation policies in the digital age”, OECD Science, Technology and Industry Policy Papers, No. 59, OECD Publishing, Paris, https://dx.doi.org/10.1787/eadd1094-en.

Hewlett-Packard (2017), “Hirotec: From smart manufacturing, to smart factory – to smart enterprise”, webpage, https://www.ptc.com (accessed 23 April 2018).

HPCI (n.d.), High Performance Computer Infrastructure website, www.hpci-office.jp/folders/english (accessed 20 April 2019).

IMEC (n.d.), “City of things”, webpage, www.imec-int.com/en/cityofthings (accessed 20 April 2019).

Kern, J. and P. Wolff (2019), “The digital transformation of the automotive supply chain – an empirical analysis with evidence from Germany and China: Case study contribution to the OECD TIP Digital and Open Innovation project”, OECD, Paris, www.innovationpolicyplatform.org/www.innovationpolicyplatform.org/system/files/imce/AutomotiveSupplyChain_GermanyChina_TIPDigitalCaseStudy2019_1/index.pdf.

Lely (2016), “Sensor technology helps farmers with even more precision and relevant information”, 17 June, www.lely.com/au/news/2016/06/17/sensor-technology/.

Lund, S., J. Manyika and K. Robinson (2016), “Managing talent in a digital age”, McKinsey Quarterly, March, McKinsey & Company, www.mckinsey.com/industries/high-tech/our-insights/managing-talent-in-a-digital-age.

OECD (2019), Digital Innovation: Seizing Policy Opportunities, OECD Publishing, Paris, https://doi.org/10.1787/a298dc87-en.

OECD (2018), “CSTP-TIP workshop: Semantic analysis for innovation policy – Summary”, OECD, Paris, www.innovationpolicyplatform.org/www.innovationpolicyplatform.org/system/files/imce/OECD%20TIP_SemanticAnalysis_summary_FINAL_0/index.pdf (accessed 13 November 2018).

OECD (2017), The Next Production Revolution: Implications for Governments and Business, OECD Publishing, Paris, https://doi.org/10.1787/9789264271036-en.

OECD (2015a), Data-Driven Innovation: Big Data for Growth and Well-Being, OECD Publishing, Paris, https://dx.doi.org/10.1787/9789264229358-en.

OECD (2015b), The Innovation Imperative: Contributing to Productivity, Growth and Well-Being, OECD Publishing, Paris, https://dx.doi.org/10.1787/9789264239814-en.

Pandolph, S. (2017), “Sephora leads the way in personalization”, Business Insider France, 22 September, www.businessinsider.fr/us/sephora-leads-the-way-in-personalization-2017-9.

Paunov, C. and S. Planes-Satorra (2019), “How are digital technologies changing innovation? Evidence from the agriculture, automotive and retail sectors”, OECD Science, Technology and Industry Policy Papers, No. 74, OECD Publishing, Paris, https://doi.org/10.1787/67bbcafe-en.

Planes-Satorra, S. and C. Paunov (2017), “Inclusive innovation policies: Lessons from international case studies”, OECD Science, Technology and Industry Working Papers, No. 2017/2, OECD Publishing, Paris, https://dx.doi.org/10.1787/a09a3a5d-en.

PSC/CAR (2017), Planning for Connected and Automated Vehicles, prepared by Public Sector Consultants and Center for Automotive Research for Greater Ann Arbor Region, www.cargroup.org (accessed 29 October 2018).

Rogers, E. (2003), Diffusion of Innovations, 5th Edition, The Free Press, New York.

RVO (2018), “Service Design-Vouchers”, webpage (in Dutch), Netherlands Enterprise Agency, www.rvo.nl/subsidies-regelingen/service-design-vouchers (accessed 23 October 2018).

The Economist/Capgemini (n.d.), “At a crossroads: Supply chains and digital transformation”, webpage, https://learnmore.economist.com/story/capgemini-supply-chains-and-digital-transformation (accessed 17 April 2018).

Tripoli, M. and J. Schmidhuber (2018), “Emerging opportunities for the application of blockchain in the agri-food industry agriculture”, Issue Paper, Food and Agriculture Organization of the United Nations and International Centre for Trade and Sustainable Development, Rome and Geneva, www.fao.org/3/ca1335en/CA1335EN.pdf.

VDA (2015), “Automation: From driver assistance systems to automated driving”, webpage, www.vda.de/en/topics/innovation-and-technology/automated-driving/automated-driving.html (accessed 29 October 2018).

Winickoff, D. (2017), “Public acceptance and emerging production technologies”, in The Next Production Revolution: Implications for Governments and Business, OECD Publishing, Paris, https://doi.org/10.1787/9789264271036-en.

Wolfert, S. et al. (2017), “Big data in smart farming – A review”, Agricultural Systems, Vol. 153, Elsevier, Amsterdam, pp. 69-80, https://doi.org/10.1016/J.AGSY.2017.01.023.

Zhu, K., K. Kraemer and S. Xu (2006), “The process of innovation assimilation by firms in different countries: A technology diffusion perspective on e-business”, Management Science, Vol. 52/10, InformsPubsOnLine, pp. 1557-1576, https://doi.org/10.1287/mnsc.1050.0487.

Note

← 1. www.oecd.org/tax/beps/.