Technology (Driving force)
The research project ‘Future scenarios of platform work’ explores the economic, labour market and societal impacts of two types of platform work – platform-determined routine work and worker-initiated moderately skilled platform work – by 2030. The project identified eight key driving forces deemed to substantially influence the development of these two types of platform work. These driving forces and associated hypotheses were used to derive potential platform work scenarios, and, from these, pointers were developed on what policy could do to make a desirable future happen and to avoid an undesirable one. This is the definition of one of the eight key driving forces identified.
The term ‘technology’ describes ‘the tools and methods used for carrying out the economic transformation process’ (Eurofound, 2018) in the digital era, such as computers and other devices capable of digitising information and processing it.
Platform work, where supply of and demand for paid labour are matched online by an internet platform, emerged due to a combination of technological developments:
- (fast) access to broadband
- mobile devices and sensors that allow connection on the move
- apps that allow for more customisation and personalisation of content and data gathering
- availability of data-gathering and analysis systems
The great innovation of platforms has been to make matching of demand and supply faster, cheaper and more convenient using digital technology. For platform work performed online, the impact has been even greater since this type of work is not location-bound. For some categories of workers with limited mobility – for example people with disabilities, older workers or people living in remote areas – this type of work offers a chance to engage in the labour market.
In both platform-determined and worker-initiated platform work (and other types of platform work too), clients post tasks on a platform, which uses a set of algorithms to match them to workers, as well as to partly manage service provision.
Several technologies are already applied to the analysed types of platform work or are likely to influence their development until 2030.
- The internet and 5G: The internet is often defined as a network of networks that allows device-to-device communication over the TCP/IP protocol (Kahn and Dennis, 2020). While this is not a new technology, its use continues to grow and its application to broaden. 5G is the fifth generation of cellular technology that is enabling ubiquitous communication and high connectivity with both individuals and objects.
- Advanced robotics: These are different from traditional robotics based on three main characteristics: they have the capability to move along more than one axis, their movement is sensor-enabled, and they can be programmed to switch between tasks. Some can also move in unstructured settings such as a factory floor where workers are also present or a public space with people.
- Autonomous vehicles: These are machines that can move around in a public space without human supervision, such as self-driving cars and self-flying drones. Their mass development is bound to 5G development due to the connectivity needs for their ‘own’ traffic management.
- Digital algorithms, data storage and data-analysis techniques: Digital data is information stored as bytes that can be read by digital devices, such as figures, images, text and logs. The more the volume of data, the more virtual space (and physical premises) is occupied. Data-analysis techniques, including artificial intelligence (AI), machine learning and deep learning, are tools and methods used to derive insights from the available data.
- The Internet of Things (IoT): This refers to the integration of physical objects with networked sensors and software. The internet serves as the channel through which information flows and objects are connected. Examples include wearable devices such as activity trackers or industrial sensors. The full capacity of this technology will depend on the adoption of 5G technologies.
- Augmented reality: This combines real-world experience with computer-generated content (Azuma, 1997; Eurofound, 2019c). Examples are smart glasses such as Google Glass and the Pokémon game, which superimpose fantasy characters upon a real setting through the screen of a smartphone. In a work setting, this could be used, for instance, in maintenance work, with the display of instructions on top of an object being repaired. Virtual reality (VR) ‘is a computer-generated scenario that simulates a real-world experience’ (Eurofound, 2019c). For platform work, this could, for example, be applied to provide training for the worker-initiated type.
- Blockchain technology: This ‘is one of the most well-known uses of distributed ledger technologies (DLT), in which the “ledger” comprises “blocks” of transactions, and it is the technology that underlies a cryptocurrency such as Bitcoin’ (Deshpande et al, 2017; Eurofound, 2019d). Distributed ledgers are considered as potentially revolutionary because they would allow for non-centralised digital verification.
- 3D printing: This is the process of manufacturing objects by adding progressive layers of material in precise locations, following a three-dimensional digitally designed model (Eurofound, 2018). This technology could, for instance, reduce the transport of small parts (relevant for the platform-determined type) and enable parts for household repairs to be printed at home (relevant for the worker-initiated type.
- Electric bikes: An electric bike or e-bike is a bicycle with an electric motor that can assist in propelling the bicycle or, alternatively, can propel it without the rider pedalling.
Internet and 5G
Platform work exists thanks to the internet. A reliable, high-speed connection is essential for platform workers to access tasks and to be either assigned to them (platform-determined platform work) or to choose them (worker-initiated platform work).
The European Commission’s Digital Economy and Society Index (DESI) measures how far digitalisation has advanced in the EU Member States (Figure 1).
Figure 1: DESI statistics for the EU27 and the UK, 2019
Source: European Commission, 2019b
Internet use both at home and on the move is increasing across the Members States. Already back in 2007, 55% of households in the EU27 and the UK had internet access; by 2017, the proportion was 89% (Eurostat, 2018). Furthermore, 67% of individuals can access the internet on the move, which is especially important for on-location platform-determined routine work, with peaks of 80% or more in Sweden, Denmark, the Netherlands, Luxembourg and Spain (Figure 2). The percentage is less than 50% in Poland (47%) and Italy (39%) (Eurostat, 2018).
Figure 2: Percentage of individuals who use a mobile phone (or smart phone) to access the internet, EU27 and the UK, 2011 and 2018
Source: Eurostat, isoc_ci_im_i
Internet availability and access to it are supported by the Member States both through investment in infrastructure and support for the development and adoption of new technologies such as 5G (European Commission, 2019d).
Over half the EU population (57%) used online services in 2017, a rise of 13 percentage points since 2012. Of course, platform work forms only part of these services, but a statistic that sheds some light on on-location platform-determined routine work is that, according to Eurostat (2018), the number of Europeans using the internet to book transportation from another individual via an app or a website was 8% in 2017, with peaks of 27% in the UK and 20% in Estonia. These figures should be read keeping in mind that not only access to technology, but also regulation, awareness, trust and user habits affect the use of these services.
Connectivity is a measure of the broadband infrastructure and of its quality. According to the DESI, in 2018, fast broadband (at least 30 Mbps) covered 83% of households, and 41% are using it already (Table 1). Ultrafast broadband (at least 100 Mbps) coverage increased to 60% of households, which is above the 50% target set by the European Commission. In terms of mobile broadband, 96% of the EU population has access to 4G mobile networks (European Commission, 2019a).
5G readiness, which is the percentage of the spectrum ready for 5G, stood at 14% in 2018. The availability of 5G depends on infrastructure development and on the implementation of efficient cybersecurity measures. At the same time, policymakers and industry need to focus on interconnectivity and interoperability (European Commission, 2019a). The recent political clashes between the United States and China about 5G technology threaten to hamper the fast deployment of this technology (Reuters, 2020).
Table 1: DESI connectivity indicators, 2019
1a1 Fixed broadband coverage
1a2 Fixed broadband take-up
1b1 4G coverage
% households (average of operators)
1b2 Mobile broadband take-up
Subscriptions per 100 people
1b3 5G readiness
Assigned spectrum as a % of total harmonised 5G spectrum
1c1 Fast broadband (NGA) coverage
1c2 Fast broadband take-up
1d1 Ultrafast broadband coverage
1d2 Ultrafast broadband take-up
1e1 Broadband price index
Score (0 to 100)
Source: European Commission, 2019a
In the EU, the price of mobile broadband was significantly cheaper in 2018 than in 2017 (European Commission, 2018c). Data, adjusted by purchasing power parity (PPP), show that there is a wide price difference between countries. The least expensive countries in 2018 were Austria, Finland, France, Italy, Luxembourg and Poland, as well as the UK. According to the Commission, ‘prices of mobile voice and data plans vary greatly across Europe. Prices went down in all consumption baskets, including 2 GB of mobile broadband and 900 voice calls or 100 voice calls went down by 37% and 44% respectively from 2016’ (European Commission, 2018a, p. 5).
Between October 2018 and March 2019 more than 95% of EU and EEA mobile phone subscribers were roaming-enabled. Almost 90% of them benefited from RLAH (roam like at home). Less than 9% were on an alternative roaming tariff (BEREC, 2019).
According to the International Federation of Robotics, sales of service robots (both robots and advanced robots) are set to increase. The sectors where increases are most likely in the next 5–10 years are medicine, logistics, agriculture, defence and public relations (Eurofound, 2019a). Sales increases are also expected, in the period 2019–2021, in sectors such as cleaning, construction, mobile platforms, inspection and underwater operations. In general, trends in the adoption of robots in services sectors are shaped by the relationship between high-investment costs and the corresponding efficiency and productivity gains. This consideration holds true for both professional and domestic use.
On-location worker-initiated moderately skilled work could be put at risk by the availability of cheap machines that can operate in domestic settings. For example, sales of smaller appliances such as vacuuming and floor-cleaning robots are increasing (although maybe not as much as estimated by the robot makers), and these could act as a cheap alternative to a human cleaner. However, it depends on whether the task involves just a floor or more complex items, as the quality the machine delivers is not necessarily the same as manual techniques.
Figure 3: Sales of service robots for personal or domestic use, estimated value, 2016 and 2017, forecast 2018–2021
Source: International Federation of Robotics
The main problem with robot vacuums and mops, as of 2020, is their inability to deal with a space where objects are left on the floor that stop the robot. For them to be useful to house-cleaning service providers, they additionally have to be able to adapt to the different spaces where they are used. Pilot projects are being run in bigger facilities such as hospitals and hotels with uneven outcomes, which depend a lot on the setting (Eurofound, 2019a).
For more skilled types of tasks, such as furniture assembly, pilot projects are also under way. It is more challenging to automate repairs, which could differ substantially.
Developments in food-processing and food-cooking appliances could provide households with an alternative to take-away or delivered meals, which could translate into fewer tasks in platform-determined work.
The market for on-location platform-determined routine work could be reduced by the adoption of autonomous vehicles in the form of self-flying or self-driving drones or self-driving cars. On-location worker-determined moderately skilled work could benefit, however, if these vehicles facilitated commuting, increasing the geographical scope of the service offered.
Many companies are running pilots for self-driving cars, including Uber and Tesla, although tests slowed down after incidents causing the death of other road users at the beginning of 2018. Trials of delivery drones are also being run by both start-ups and established companies; in the UK, Amazon is testing drone deliveries of small parcels (maximum of a few kilograms) to customers in rural areas to minimise the risk of damage (Eurofound, 2019b). Deliveries by drones on wheels are also being tested, and the annual number of patents registered for ‘last mile’ delivery drones is high and has increased since 2010 (from 1,550 to 2,500 per year) (Eurofound, 2019b).
Although experiments in vehicle and drone automation are being conducted to various degrees in different countries and by different companies, a common view among experts in the sector is that ‘fully autonomous vehicles are not expected to be deployed until the mid-2030s’ (Eurofound, 2019b). This slow adoption essentially stems from the same challenge that robotics applications have to overcome – the difficulty of navigating through unstructured and changing environments. Challenges include harm to people and damage to property; regulatory and insurance guidelines; noise levels in residential areas and adverse weather conditions (for flying drones). Autonomous vehicles also need a solid sensor infrastructure that can support their data navigation needs in terms of fast connectivity and fast processing power, which should eventually materialise with 5G. Last but not least, consumers’ trust in this type of technology will also be needed for it to be widely adopted.
Digital algorithms, data storage and data-analysis techniques
Moore’s Law states that processing power doubles every year; with time and improvements in microprocessors, this speed has slowed down to 18 months despite miniaturisation efforts and the addition of cores to computing machines. Experts forecast that by 2030 Moore’s Law will no longer hold (Theis and Wong, 2017). Still, a simple example that illustrates the growth of data-processing speed is the fact that smartphones can perform much better than the first calculator or even a normal household computer 10 years ago. Processing speed goes hand in hand with data production. The vast volume of data produced by devices and machinery prompted scholars to call this explosion of data the ‘Zettabyte Age’, a zettabyte being 270 bytes (Kitchin, 2014). In 2018, 90% of data in the world had been created in the previous two years (Forbes, 2018), and this trend is growing as more and more devices and sensors are used.
In 2017, The Economist deemed big data as the ‘new oil’, highlighting that great knowledge and power comes from hoarding and analysing data to gain marketing and business insights. Increasingly cheap data storage and processing software make it easy to collect and analyse huge amounts of data. The more high-quality and correctly tagged data available, the more it is possible to automate processes and apply machine-learning techniques, which could improve the algorithmic decision-making of the platform.
All platform matching processes are digital, and the digitisation of information gives a wealth of data to the platform. Workers’ locations, the number of transactions and customer ratings are all inputs into understanding the business and offering customers what they are looking for.
One aspect of this that needs to be considered is the transparency for the worker and the client regarding the data collected, how they are stored and what they are used for. If the platform uses the data to achieve productivity gains or even sells them to third parties for unrelated use, discussions on the distribution of the revenue might emerge. If algorithms and the data collected influence task assignment to workers, questions of fairness, objectivity and possibilities for redress also arise.
Internet of Things (IoT)
IoT applications, including sensors attached to objects and wearables, are relevant to the tasks performed in on-location platform-determined routine work; for instance, sensors can be used to track the location and movements of delivery workers or of a car.
Eurofound’s (2019e) study of wearables noted that:
Forecasts suggest that the market for IoT will continue to grow considerably. Bosche et al (2018), for instance, estimate that the combined markets of the IoT will grow to about USD 520 (€450.75) billion globally by 2021, more than double the USD 235 (€203.75) billion spent in 2017.
Interoperability, interconnectivity and cybersecurity are market prerequisites for the rapid adoption and commercialisation of IoT, but as of early 2020 not yet fully implemented. On 29 January 2020, the European Commission approved a 5G Toolbox, setting out guidelines for secure deployment of the technology and for mitigating risks of interference from a third country via the 5G supply chain.
Servitisation is the process through which machinery and objects are monitored at a distance by manufacturers. Remote monitoring using sensors can raise alerts about small faults and can predict failures, enabling the manufacturer to send a technician to fix the item as part of a service agreement. Depending on the level of IoT adoption, for some of the tasks associated with on-location worker-initiated moderately skilled work, this technology could present two possible options: the intervention of platform workers is superseded by that of a technician, or platform workers use tools and devices that can help them diagnose the fault.
For both on-location platform-determined routine work and on-location worker-initiated moderately skilled work, the potential use of sensors for control and surveillance purposes raises questions linked to data protection, privacy and labour rights.
Augmented reality could contribute to making on-location platform-determined routine work safer by, for example, providing road safety information in an easily and comfortably accessible way to delivery drivers.
Handymen doing on-location worker-initiated moderately skilled work could benefit from the use of smart glasses or enabled smartphones, which could give them extra information about the make and model of items that they have to fix. Furthermore, virtual reality (VR) could be used for training purposes, making respective tasks better accessible for lower-skilled workers.
The use of smart glasses for entertainment and everyday use has failed in the past, but experiments are being carried out with repurposed models to explore whether they can be used in industrial work settings for technical guidance in repairing equipment. As of early 2020, the investment costs in both hardware and software are an issue. If this technology becomes cheap to a point that it is better than a do-it-yourself video, it could be appealing to some customers.
Blockchain, and other distributed ledger technologies, are still at a very early stage of adoption, except for the cryptocurrency market. Several pilots are being run in different areas, including experiments by well-known banks in managing transactions, smart contracts, certification of education degrees and vote-casting (Eurofound, 2019d). Since the technology is still in its infancy, there are no statistics on usage, but there are 660 use cases in the EU as listed in the EU Blockchain Observatory and Forum at the time of writing. There is potential for blockchain to be used as an underlying platform for financial and assets exchanges but before this happens, blockchain tests must be successful and standards must be agreed to guarantee interoperability.
Electric bikes and scooters are already in use, and this could increase over time. In 2018, the sales of e-bikes surpassed that of traditional bikes in the Netherlands (Bovag, 2019). The European Cyclists’ Federation, an employer federation, estimates that 30 million electric bikes would be sold between 2019 and 2030 (ECF, 2019); this figure has been recently adjusted upward (Forbes, 2019).
An increase in the spread of e-bikes can be beneficial for (potential) platform-determined workers as they reduce the physical requirements to conduct the tasks, and hence make them better accessible for specific groups in the population (such as older people or those with disabilities).
An intangible asset for fast technology adoption is the level of digital skills required to operate these systems, both from the point of view of the platform and the workers. If new systems require a high level of specialist skills, the labour market might not be able to immediately fill the gap; if the systems are easy to use, this might not be an obstacle.
According to the DESI, 43% of the EU population still lack basic digital skills. On the positive side, 81% of Europeans go online regularly (at least once per week), up by 2 percentage points compared with 2016 (European Commission, 2018b). The number of science, technology, engineering and mathematics (STEM) graduates is also rising gradually: 19.1 graduates per 1,000 people aged 20–29 years in 2015, compared to 18.4 in 2013.
The project applied a foresight methodology to derive possible future scenarios. As part of this, at least two realistic and mutually exclusive hypotheses were drafted for each of the key drivers to depict their potential future development.
Three hypotheses were identified for this driver.
1. Fast adoption
Fast adoption of technology might occur if it is low cost, if people have a sufficient level of skills and trust in the technology, and if supportive framework conditions exist (such as legislation, public funding, infrastructure, energy consumption). For example:
- 5G connectivity for the vast majority of the EU population
- sensors widely deployed to track and accompany transport services in platform-determined tasks
- autonomous shuttle services and use of autonomous vehicles for deliveries of consumer goods in many cities
- advanced robotics able to replace a large share of standard household services (like cleaning and gardening)
This hypothesis links with the following scenarios:
- Platform work scenario: Technological dystopia
- Platform work scenario: Robust and consistent response by the EU
- Platform work scenario: Regulate to innovate
- Platform work scenario: Worker and trade union power
2. Slow adoption
Slow adoption of technology might occur if it is high in cost, if people’s skills and trust in the technology remain low, and if there is an absence of suitable framework conditions. For example:
- citizens do not trust autonomous shuttle services in cities, or the offer is limited to small areas only, decreasing the attractiveness for users
- household robots are limited to basic services still requiring human input for more complex tasks
- technological advancement influences platform work, such as requiring higher skills in the worker-initiated type (as basic tasks are replaced by the machines), or subjecting the worker-initiated type more to automated surveillance (for example, through AI), aligning it more to the characteristics of the platform-determined type
3. Moderate adoption
The pace of adoption is likely to be moderate if some problems of connectivity, cost, and trust or usage exist. For example:
- 5G available just in big cities, with uneven connectivity across regions, resulting in selected use rather than more general use
- robotics develops more than in H2, but people refuse distribution of their data without individual approval
This hypothesis links with the following scenarios:
- Platform work scenario: Business as usual
- Platform work scenario: Worst case
- Platform work scenario: Clear employment status benefits all workers and worker cooperatives
- Platform work scenario: Regulation-driven and emphasis on public platforms
- Platform work scenario: Matching, only matching
- Platform work scenario: Economic downturn and partial regulatory response
- Azuma, R.T. (1997), ‘A survey of augmented reality’, Presence: Teleoperators and Virtual Environments, Vol. 6, No. 4, pp. 355–385.
- BEREC (2019), International roaming: BEREC benchmark data report, October 2018 – March 2019, Riga.
- Bovag (2019), Zonnig 2018 stuwt omzet fietsbranche naar record, Bunnik.
- ECF (European Cyclists’ Federation) (2019), E-bike sales can reach 30 million unit per year by 2030, Brussels.
- Economist (2017), ‘The world’s most valuable resource is no longer oil, but data’, May 6.
- Deshpande, A., Stewart, K., Lepetit, L. and Gunashekar, S. (2017), Distributed ledger technologies/blockchain: Challenges, opportunities and the prospects for standards, British Standards Institution, London.
- Eurofound (2018), Automation, digitisation and platforms: Implications for work and employment, Publications Office of the European Union, Luxembourg.
- Eurofound (2019a), Digital age: Advanced robotics: Implications of game-changing technologies in the services sector in Europe, Dublin.
- Eurofound (2019b), Digital age: Autonomous vehicles: Implications of game-changing technologies in the services sector in Europe, Dublin.
- Eurofound (2019c), Digital age: Virtual and augmented reality: Implications of game-changing technologies in the services sector in Europe, Dublin.
- Eurofound (2019d), Digital age: Blockchain: Implications of game-changing technologies in the services sector in Europe, Dublin.
- Eurofound (2019e), Digital age: Wearable devices: Implications of game-changing technologies in the services sector in Europe, Dublin.
- European Commission (2018a), Commission staff working document: Digital Economy and Society Index (DESI) 2018, SWD(2018) 198 final part 2/6, Brussels.
- European Commission (2018b), Digital economy and society index report 2018, Brussels.
- European Commission (2018c), Mobile Broadband Prices in Europe 2018, Publications Office of the European Union, Luxembourg.
- European Commission (2019a), Digital Economy and Society Index report 2019, Connectivity, Brussels.
- European Commission (2019b), Digital Economy and Society Index report 2019, Human capital, Brussels.
- European Commission (2019c), EU Blockchain Observatory and Forum, accessed 19 June 2020.
- European Commission (2019d), Towards 5G , Brussels.
- European Commission (2020), Secure 5G networks: Questions and answers on the EU toolbox, web page, accessed 10 March 2020.
- Eurostat (2018), Digital economy and society statistics - households and individuals.
- Forbes (2018), How much data do we create every day? The mind-blowing stats everyone should read, 21 May.
- Forbes (2019), 100 million extra e-bike purchases by 2030, graphs NGO using 2018's stellar EU sales figures, web page, accessed 10 March.
- International Federation of Robotics (IFR) (undated), statistics, web page, accessed 19 June 2020.
- Kahn, R. and Dennis M. (2020), Internet, Encyclopaedia Britannica, web page, accessed 19 June 2020.
- Kitchin, R. (2014), ‘ Big Data, new epistemologies and paradigm shifts', Big Data & Society, April–June: 1–12.
- Reuters (2020), ‘ U.S. urges EU to use 5G by Ericsson, Nokia, Samsung, seen on par with Huawei’, 19 February.
- Theis, T. and Wong, P. (2017), ‘The end of Moore's Law: A new beginning for information technology’, Computing in Science & Engineering, Vol. 19, No. 2, pp. 41–50.