Algorithmic management in the logistics sector in France

Bisaschi, Luca; Calderoni, Paolo; Garces, Inazio; Lechardoy, Lucie; Nardoni, Sara Working Paper Algorithmic management in the logistics sector in France JRC Working Papers Series on Labour, Education and Technology, No. 2025/04 Provided in Cooperation with: Joint Research Centre (JRC), European Commission Suggested Citation: Bisaschi, Luca; Calderoni, Paolo; Garces, Inazio; Lechardoy, Lucie; Nardoni, Sara (2025) : Algorithmic management in the logistics sector in France, JRC Working Papers Series on Labour, Education and Technology, No. 2025/04, European Commission, Joint Research Centre (JRC), Seville This Version is available at: https://hdl.handle.net/10419/322091 Standard-Nutzungsbedingungen: Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Zwecken und zum Privatgebrauch gespeichert und kopiert werden. Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich machen, vertreiben oder anderweitig nutzen. Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, gelten abweichend von diesen Nutzungsbedingungen die in der dort genannten Lizenz gewährten Nutzungsrechte. Terms of use: Documents in EconStor may be saved and copied for your personal and scholarly purposes. You are not to copy documents for public or commercial purposes, to exhibit the documents publicly, to make them publicly available on the internet, or to distribute or otherwise use the documents in public. If the documents have been made available under an Open Content Licence (especially Creative Commons Licences), you may exercise further usage rights as specified in the indicated licence. https://creativecommons.org/licenses/by/4.0/ Centre Algorithmic Management in the Logistics Sector in France JRC Working Papers Series on Labour, Education and Technology 2025/04 L. Bisaschi, P. Calderoni, I. Garces, L. Lechardoy, S. Nardoni This publication is part of a working paper series on Labour, Education and Technology by the Joint Research Centre (JRC). The JRC is the European Commission’s science and knowledge service. It aims to provide evidence-based scientific support to the European policymaking process. The contents of this publication do not necessarily reflect the position or opinion of the European Commission. Neither the European Commission nor any person acting on behalf of the Commission is responsible for the use that might be made of this publication. For information on the methodology and quality underlying the data used in this publication for which the source is neither Eurostat nor other Commission services, users should contact the referenced source. The designations employed and the presentation of material on the maps do not imply the expression of any opinion whatsoever on the part of the European Union concerning the legal status of any country, territory, city or area or of its authorities, or concerning the delimitation of its frontiers or boundaries. Contact information Name: Ignacio González Vázquez Address: Joint Research Centre, European Commission (Seville, Spain) Email: ignacio.gonzalez-vazquez @ec.europa.eu EU Science Hub https://joint-research-centre.ec.europa.eu JRC142292 Seville: European Commission, 2025 © European Union, 2025 The reuse policy of the European Commission is implemented by the Commission Decision 2011/833/EU of 12 December 2011 on the reuse of Commission documents (OJ L 330, 14.12.2011, p. 39). Except otherwise noted, the reuse of this document is authorised under the Creative Commons Attribution 4.0 International (CC BY 4.0) licence (https://creativecommons.org/licenses/by/4.0/). This means that reuse is allowed provided appropriate credit is given and any changes are indicated. For any use or reproduction of photos or other material that is not owned by the EU, permission must be sought directly from the copyright holders. All content © European Union 2025 How to cite this report: Bisaschi, L., Calderoni, P., Garces, I., Lechardoy, L. and Nardoni, S., Algorithmic Management in the Logistic Sector in France, European Commission, Seville, 2025, JRC142292. 1 Contents Abstract ..................................................................................................................................................................................... 2 Executive summary ............................................................................................................................................................ 3 1 Introduction ................................................................................................................................................................... 4 2 Methodology ................................................................................................................................................................. 5 3 Applications of algorithmic management tools in the logistics sector in France and its possible impacts ................................................................................................................................................................... 6 3.1 French strategy for Artificial Intelligence ....................................................................................... 6 3.2 Algorithmic management tools in the logistics sector ........................................................... 7 4 Business model and work organisation ........................................................................................................ 9 4.1 Business model and the delivery of services ............................................................................... 9 4.2 Organisation and coordination of work processes ................................................................. 14 4.3 Occupation, tasks and skills ................................................................................................................ 15 5 Working conditions ................................................................................................................................................. 16 5.1 Work intensification ................................................................................................................................. 16 5.2 Autonomy ...................................................................................................................................................... 17 5.3 Social interaction ....................................................................................................................................... 17 6 Ethics of AI .................................................................................................................................................................. 17 7 Surveillance and data privacy issues in the era of algorithmic management .................... 18 8 Conclusions ................................................................................................................................................................. 24 References ............................................................................................................................................................................ 26 List of figures...................................................................................................................................................................... 28 List of boxes ........................................................................................................................................................................ 29 Appendix ................................................................................................................................................................................ 30 1. Distribution of interviews ........................................................................................................................... 30 2 Abstract This working paper provides an assessment of the impacts of algorithmic management (AM) tools on working conditions, job quality, and industrial relations in the logistics sector of France. It analyses the different ways algorithmic management technologies affect coordination, distribution and content of tasks, employee autonomy, job profiles, and industrial relations in logistics companies. The report presents a comprehensive study conducted through desk research, stakeholder consultation, and interviews with French start-ups, industry associations, public regulatory bodies, academia, and large companies. The French Strategy for Artificial Intelligence and the recent expansion of AI in French companies are discussed, along with some key applications of algorithmic tools. While algorithmic management tools offer several benefits for logistics companies, including optimizing operations and enhancing customer satisfaction, there are also concerns about the ethical and legal implications of rising technologies, particularly regarding worker privacy and the potential for bias or discrimination. Recent regulations and guidelines are presented to ensure the fair and transparent use of these tools. The paper concludes by discussing the need to balance the benefits of algorithmic tools with the potential risks and ethical concerns, and the role of the market and public policy in the coordination of economic activities in the logistics sector. Keywords: Algorithmic Management, Logistics, Working Conditions, Job Quality Joint Research Centre reference number: JRC142292 3 Executive summary The report presents a selection of case studies of algorithmic management technologies (AMT) in the logistics sector in France. Specifically, the case studies examine the use of various AMTs, including AI tools, which are increasingly deployed to manage workers in the sector. France is home to some of the world's major logistics players, including shipping companies, parcel deliveries, freights deliveries, or terminal operators, and many of the largest logistics companies in the world operate on French soil. French governments have tried to foster the development of new start-ups, some of which operate in the development of AMTs and AI technologies to manage physical and informational flows. The logistics industry is a significant contributor to the French economy, accounting for 10% of GDP, 150,000 firms, and 1.8 million jobs. The French government has invested over 1.5 billion euros in AI development, and private co-financing has added an additional 500 million euros. While algorithmic management tools have benefited the logistics industry, it has also created questions about ethical applications and technological use in data protection and privacy. Interviewees explained that algorithms have enabled machines to sense, comprehend, learn, and act at human-like levels, raising concerns about their implications for employees, such as loss of autonomy, increased surveillance, and biased decision-making. Moreover, workers, especially platform workers, have mobilised against algorithmic management, contesting its impact on their working conditions, as France has historically been a land for workers' mobilisation in the logistics sector. For example, French civil society has engaged with non-profit experiments in the urban biking delivery sector, and trade unions have opened a debate on the impact of new digital technologies on working conditions. The stakeholder consultation presented in this report suggests that the increasing adoption of AMTs in the logistics sector in France has both positive and negative impacts. While it has enabled companies to automate and streamline processes, predict demand, and reduce operational costs, it has also created questions about ethical applications and technological use in data protection and privacy. This has raised concerns about AMTs’ impact on workers' autonomy, increased surveillance, and biased decision-making. Therefore, stakeholders should consider the implications of AMT on working conditions in the logistics sector, and design measures to mitigate the negative impacts of algorithmic management. Main findings The findings reveal that AMT adoption has become widespread in the French logistics industry in recent years, offering competitive advantages to companies that invest in it. Many logistics companies have invested in AMT to automate and streamline processes and to predict and forecast demand, resulting in lower operational costs and higher revenues. For example, Smart Warehouse Systems are used to recognise patterns and dependencies from unstructured data using IoT, AI, and cloud computing. These appliances can adapt independently and dynamically to new circumstances throughout the entire logistics system, making monotonous jobs simpler, and operations more efficient and cost-effective. However, this has also made certain warehouse jobs redundant and increased the importance of engineering and managerial jobs. Quick guide The first part of this report introduces the topic and its relevance. Section 2 describes the methodology followed throughout the study, while Section 3 presents a description of the logistics sector in France and the technologies used there. Section 4 lays the groundwork for the evidence collected throughout the stakeholder consultations with first-hand data and responses on the different impacts of algorithmic management tools on business model and the delivery of services, Section 5 on working conditions, Section 6 on the ethics of AI, and Section 7 on surveillance and data privacy issues. Finally, Section 8 concludes the report with a recapitulation of the data collected and its different impacts. 4 1 Introduction The rapid advancement of the digital revolution and the growing connectivity in the workplace are reshaping the world of work. Digital tools and algorithmic management practices are now being adopted in traditional workplaces as mechanisms for work coordination. The digitalisation of work, along with digital monitoring and algorithmic management – defined as the use of computerprogrammed processes to coordinate labour within an organisation – creates new business opportunities, enhances efficiency, and optimizes workflows. However, this shift also raises concerns about working conditions, the potential decline in job quality, and the increased risk of worker surveillance 1 . Over the past few decades, the use of algorithms across industries has quickly spread throughout the world. Especially impacting the most industrialised countries, algorithms have transformed our societies, how companies and markets operate, and how humans interact with machines: from academia to business practices, and from operations to human management. Algorithmic management refers to the use of AI and other algorithmic techniques to manage and supervise workers and business processes. Essentially, it involves automating certain managerial tasks that would normally be carried out by human managers. In practice, algorithmic management technologies (AMT) can take many different forms. For example, they can involve using machine learning (ML) algorithms to analyse employee performance data and make decisions about promotions, bonuses, and other forms of compensation. They can also involve using chatbots or other automated tools to communicate with employees and answer their questions or using predictive analytics to forecast future business needs and adjust staffing levels accordingly. Digital labour platforms have been pioneers in adopting surveillance technologies and algorithmic management practices 2 , 3 . This has been facilitated both by their inherently digital operating environment and by the informal nature of their working relationships, which has often allowed them to bypass many of the restrictions typically found in traditional workplaces. However, algorithmic management and digital surveillance are now increasingly permeating conventional work settings as well, prompting broader concerns about their impact on labour processes, the organisation of work, and the balance of power in the workplace 4 . For example, there are concerns about algorithmic bias, where algorithms may inadvertently discriminate against certain groups of workers or perpetuate existing inequalities. There are also concerns about worker privacy and autonomy, as well as potential legal liability if algorithmic decisions lead to negative outcomes for workers. Therefore, it is important for organisations to carefully consider the implications of algorithmic management and ensure that they have robust policies and procedures in place to address these issues. The case studies presented in this report were conducted at the end of 2022 in the frame of a broader research study commissioned by the Joint Research Centre (JRC) to Open Evidence, investigating the impact of algorithmic management tools (AMT) on work organisation. The research study was part of a broader project carried out jointly by the JRC and the International Labour Organisation (ILO) focusing on country-sector analyses in the logistics and healthcare 1 Rani, U., Pesole, A. and Gonzalez Vazquez, I., Algorithmic Management practices in regular workplaces: case studies in logistics and healthcare, Publications Office of the European Union, Luxembourg, 2024, doi:10.2760/712475, JRC136063. 2 Pesole, A., Urzì Brancati, M.C., Fernandez Macias, E., Biagi, F. and Gonzalez Vazquez, I. 2018. Platform Workers in Europe Evidence from the COLLEEM Survey, EUR 29275 EN, Publications Office of the European Union, Luxembourg, 2018, ISBN 978-92-79-87996-8, doi:10.2760/742789, JRC112157. 3 ILO. 2021. World Employment and Social Outlook 2021: The Role of Digital Labour Platforms in Transforming the World of Work. International Labour Organization. 4 Baiocco, S., Fernández-Macías, E., Rani, U. and Pesole, A., 2022. The Algorithmic Management of work and its implications in different contexts, Seville: European Commission, JRC129749. 5 sectors in two European countries (France and Italy) and two non-EU countries (India and South Africa) allowing for a comparative analysis across countries with different levels of development 5 . The aim of the overall project was to assess applications of algorithms in the workplace and examine their impacts on the coordination of work and working conditions, together with the role of social dialogue in the adoption and implementation of these technologies. The logistics sector was chosen as an example of a highly digitalised sector in which the use of digital monitoring and algorithmic management technologies has been widely documented in the literature. In contrast, in the healthcare sector the use of algorithmic management as part of digital health platforms is more recent and less well documented. This report focuses on the logistics sector in France. 2 Methodology Each case study considered in the project was conceived to include a number of qualitative, indepth, semi-structured interviews and field visits in an establishment using algorithmic management for the coordination of work processes, to be complemented by desk research as appropriate 6 . Specifically, the case studies involved interviews with individuals in different functions and roles in each establishment including managers at different levels, union and worker representatives, technology specialists, and workers affected by the technology. A semi-structured interview guide was designed and organised around a set of core themes and questions intended to investigate the impact of algorithms and digital solutions on work organisation and working conditions. All interviews were recorded and later transcribed. A content analysis of each interview was carried out: interviews’ transcripts were individually studied and developed from generalised contents towards robust analyses progressively crystalising on the key dimensions covered. The main topics covered in the interviews include: (i) impacts on business model and the delivery of services; (ii) impacts on organisation and coordination of work processes; (iii) impacts on occupations, tasks and skills; (iv) impacts on work intensification; (v) impacts on autonomy; (vi), impacts on social interaction; (vii) ethical concerns surrounding the use of AI; (viii) impacts on monitoring and surveillance; and (ix) impacts on data privacy issues. The participation and involvement of different stakeholders in the research allows to provide a balanced understanding of the implementation of the digital solution, the level of algorithmisation of the technology in place, and its impacts in the workplace. In addition to the qualitative interviews, the case studies are supplemented with information and documentation provided by the interviewees and other working papers. Outreach activities for the selection of the establishments to be used as case studies in the French logistics sector involved accessing logistics companies of all sizes across France, including multinationals and start-ups. However, from the beginning of the project, French logistics companies were particularly reluctant to participate in the study. The main reasons for companies declining participation as a case study at the establishment-level included concerns on the legislative process of the EU, time constraints, lack of interest, and unavailability due to lack of resources. Moreover, the topic of algorithmic management in the logistics sector has become particularly sensitive in France due to recent scandals and ongoing worker protests. Concerns over precarious working conditions, surveillance, and the perceived lack of transparency in algorithmic decision-making have fueled criticism. As a result, companies may be reluctant to expose themselves to public scrutiny or regulatory attention, making them hesitant to participate in the study. In total, around 50 different companies using algorithmic management tools were contacted to participate as a case study at the establishment-level. Consequently, the methodology of the research study for the French logistics was modified to overcome the concerns raised by companies and to collect sufficient data whilst maintaining the overarching scope of the research project. In 5 Rani, U., Pesole, A. and Gonzalez Vazquez, I., Algorithmic Management practices in regular workplaces: case studies in logistics and healthcare, Publications Office of the European Union, Luxembourg, 2024, doi:10.2760/712475, JRC136063. 6 Ibid. 6 the end, a total of 25 stakeholders from diverse backgrounds in logistics were interviewed across 14 different organisations or enterprises. This enabled to gain a better understanding of the impact of algorithmic management tools in the workplace. Stakeholders included French start-ups and large logistics companies, industry associations, academia and public regulatory bodies. All interviews, desk research, data collection and reporting were carried out between December and March 2023. Interviews were conducted online via MS Teams with a pre-established questionnaire sent in advance (see appendix for more details). 3 Applications of algorithmic management tools in the logistics sector in France and its possible impacts 3.1 French strategy for Artificial Intelligence According to the French government, the logistics industry in France accounts for 10% of the GDP, 150,000 firms, and 1,8 million jobs (four times as much as the automobile industry) 7 . Moreover, between the years 2018-2023, the French government has dedicated more than 1.5 billion euros to the development of AI in the country 8 , and an additional 500 million euros coming from private cofinancing has been poured into the tech industry 9 . The French National Strategy on AI (Stratégie nationale pour l'intelligence artificielle), labelled ‘AI for Humanity’, was launched in 2018 and aims at implementing the recommendations of Cédric Villani’s report on AI launched that same year (Villani et al., 2018). The report builds on four axes 10 . Axis 1: To enhance the French research ecosystem on AI through additional financial resources, new infrastructures, and more links with the industry. Axis 2: To accelerate AI dissemination in the economy, public administration, and society. Axis 3: To encourage the development of a ‘Trustworthy AI’, both at the national and international levels. Axis 4: To develop and reinforce AI educational programmes to increase AI content across universities. All stakeholders interviewed agreed that the National Strategy has played an important role in the adoption of new AI tools in the industry, as well as algorithmic management technologies across different sectors. Through the aid of state financing and the pooling together of talent, interviewees agreed that a new push in the use of AI in small and medium enterprises was growing, including new AI tools being invented in France. The Villani report highlights the need to build a data-focused economic policy to catapult Europe into becoming a world leader in AI. The report suggests that France and Europe must design a tailored model to capitalize on high protection standards through the European General Data Protection Regulation (GDPR) 11 . To achieve this, public authorities must introduce new ways of producing, sharing and governing data by making data a common good. The State is a key driver in these various areas of transformation, and public authorities must ensure that they adopt the necessary material and human resources to factor AI into the way they address public policy, with the aim of both pursuing modernization and acting as an example to be followed (Villani et al., 2018). Additionally, the report suggests setting up shared sector platforms that provide secure and tailored access to different participants in these ecosystems, including researchers, companies, and public authorities. Moreover, these shared platforms would enable sharing of useful data for the development of AI, as well as foster resources and extensive computing infrastructure (Villani et al., 7 Ministère de la Transition écologique et de la Cohésion des territoires (2021). https://www.ecologie.gouv.fr/logistique-en france#:~:text=La%20logistique%20est%20%C3%A0%20ce,%2C8%20million%20d'emplois 8 France AI Strategy Report (2021). https://ai-watch.ec.europa.eu/countries/france/france-ai-strategy-report_en 9 French AI Strategy (2018). https://www.euractiv.com/section/digital/news/french-ai-strategy-tech-sector-to-receive-overe2-bln-in-next-5-years/ 10 OECD.AI Policy Observatory (2021). https://oecd.ai/en/dashboards/policy-initiatives/http:%2F%2Faipo.oecd.org%2F2021data-policyInitiatives-25374 11 AI For Humanity (2018). https://www.aiforhumanity.fr/ 13 — Routing and scheduling: With the use of algorithms, logistics companies can optimize routes and schedules to improve delivery times and reduce costs. These tools can analyse data on traffic patterns, weather conditions, and delivery destinations to determine the most efficient routes. — Inventory management: Algorithms can monitor inventory levels in real-time and make recommendations on when and how much to reorder. These tools can help logistics companies avoid stockouts and minimize waste. — Quality control: Algorithms can be used to monitor and analyse data from sensors and other sources to identify quality control issues in real-time. This can help logistics companies identify and address problems before they lead to customer complaints or product recalls. — Predictive maintenance: Algorithms can also analyse data from sensors and other sources to predict when equipment is likely to fail. These tools help logistics companies schedule maintenance proactively, reducing downtime and extending the lifespan of equipment. — Customer service: Finally, algorithms can be used to analyse data on customer preferences and behaviours as to personalize customer service interactions. This can involve recommending products or services, providing tailored offers or discounts, and responding to customer inquiries in real-time. Box 2. Case study on French Logistics 3 A major programme within the Foundation is a last-mile logistics platform for the optimisation of delivery flows, traffic prediction, and mobility transition of delivery vehicle fleets. One algorithmic management tool within platform allows management of the organisation of delivery areas by adapting its needs and developing the roadway according to the requirements of drivers. Thanks to the tool, drivers have easy access to important information before and throughout deriving goods and when planning future routes. This includes information on road traffic and delivery areas, the number and location of delivery areas, the availability of delivery areas, the typology of parked vehicles, the duration of parking zones, and the nature of parking areas. The tool shares similarities with another algorithm used by Amazon as a last-mile solution for faster delivery, lower costs, and a better customer experience – quickly and easily managing its logistics in a few clicks 27 . In essence, the tool developed by French Logistics 3 enables digitalising commercial vehicles’ disc parking in France (a system of allowing time-restricted free parking through the display of a parking disc showing the time at which the vehicle was parked). The first step in using the tool is a simplified parking declaration, with information on parking time, location, and alerts, via selfgeolocation of delivery spaces. The second step is to monitor and inform the driver to know where they can park, with geolocation of delivery areas and the availability level for each; broken down between free, available for 5mins and unavailable. The third step is to ‘consult and adapt’, with drivers being able to reserve parking spaces and delivery areas in advance. Figure 1 illustrates the tool’s dashboard. On the left-hand side, we see the mobile interface showing the parking time limit, location, and other key information for drivers, and on the right-hand side, we see a heatmap with parking availability broken down into the three aforementioned levels. Figure 1: Example of dashboard 27 AWS (2023). https://aws.amazon.com/blogs/supply-chain/aws-last-mile-solution-for-faster-delivery-lower-costs-and-abetter-customer-experience/ 14 Source: Material provided by French Logistics 3 The operating principles of the tool are simplicity and confidentiality, with a declaration of parking under a simple click, a geolocation only at the time of parking (inactive when the vehicle is moving to limit data privacy issues), and phone alerts on the remaining parking time. Moreover, the algorithm allows for managing the evolution of the driver’s needs by simulating the impacts on traffic. The algorithm predicts traffic and free parking spaces using data from public cameras and sensors to recognise cars, motorbikes, and bicycles on the street. By using this tool, companies can centralise the different delivery routes of vehicle fleets and easily manage their logistics from one space. With live updates of the delivery status, position and parking time, companies can manage the available resources at all times and better organise themselves around peak activities. Using the tool helps increase flow optimisation and increase route efficiency. 4.2 Organisation and coordination of work processes Algorithmic management tools can change the way work is organized and managed. Managers may rely more heavily on data and analytics to make decisions about how work is assigned and performed. This can result in workers being assigned tasks based on data-driven metrics, such as productivity or efficiency, rather than their own preferences or skills. For example, in a call centre, an algorithmic management tool may assign calls to workers based on their previous call performance, rather than allowing workers to choose which calls they take based on their own preferences or skills. One example of how algorithmic management tools impact work organization in the French logistics sector is through the use of route optimization algorithms for delivery drivers. These algorithms use real-time traffic data to calculate the most efficient routes for drivers to take to complete their deliveries, reducing travel time and increasing productivity. Another example is the use of predictive analytics tools to forecast demand and optimize inventory levels in warehouses. This can help to reduce overstocking and understocking of goods, which can improve efficiency and reduce waste. Additionally, the use of algorithmic management tools may require new skills and training for workers, particularly in the use of data analytics and technology. This can lead to changes in job roles and responsibilities and may require additional investment in training and development programs. To collect information on the impacts of AMTs on work coordination, organisation and workflows, French AI associations and non-profit technology organisations were interviewed. The co-founder of a French AI association founded following the Villani Report that helps start-ups raise funds, share ideas, and pool talent together. French logistics association has several working groups on AI, including one on HR focusing on the reorganisation of teams, the replacement of humans, and the questions of reassigning employees. Other working groups involve the transport and delivery sector, supply chain, stock management, flow optimisation, and reducing dangers in warehouses. The 15 founder explained that the objective is to have French digital sovereignty to not depend on other countries and be able to compete with Silicon Valley. For the founder, the goal of AI applications is generally to lower costs, decrease the time needed to complete tasks and increase the frequency, reactivity, and profitability of teams. Several case studies within the association include new standards on the respectful use of drones in the industry. The interest is based on the increasing use of drones in warehouse storage areas to coordinate employees and manage workers in factories. Likewise, drones are used for safety, flow optimisation, and risk reduction. Another example discussed was the increasing use of drones to transport goods between remote areas and islands around the world. Whilst before human-piloted helicopters were widely employed to transport essential goods in remote areas, such as medicine across short distances and between islands, nowadays drones are being used for deliveries. One example discussed with the association was the use of drones in rural Canada to deliver medical supplies to remote areas and indigenous communities 28 . Many technology experts realised that they could automate logistics by sending unmanned drones that are more fuel efficient, faster, and more environmentally friendly, said the interviewee. Furthermore, he stated that drones were being used across French companies for airlifts - location, air capture, radar, camera etc. Another French supply chain association interviewed explained that AI management tools were mainly being used in Warehouse Management Systems (WMS) and Transport Management Systems (TMS). Across the transport and warehouse applications, digital technologies are rapidly expanding to manage employees and increase work organization, the interviewee said. 4.3 Occupation, tasks and skills Algorithmic management tools can require workers to develop new technical skills to use and interact with the algorithms. Workers may need to be trained in how to use digital devices or how to interpret the data they generate. Often, these two tasks are assigned to different categories of workers: low-rank workers mainly produce information, while medium-high rank workers mainly interpret and manipulate information. 28 The University of British Columbia (2022). https://beyond.ubc.ca/forward-happens-here-rural-health/ 16 Figure 2: Task distribution and job hierarchisation according to margins of data manipulation at Amazon.com warehouses Source: Massimo 2020a In addition, workers may need to be comfortable with technology and data analysis in order to effectively use the algorithmic tools. For example, a customer service representative may need to learn how to use a chatbot to respond to customer inquiries, or how to analyse customer data to identify patterns and trends. Some examples of how AMT are impacting skills in the French Logistics sector include: — Technical skills: The use of AMT requires employees to have technical skills to operate and interpret the data generated by these tools. For example, workers need to be proficient in using warehouse management systems, inventory management software, and other similar tools to manage and optimize their work. — Analytical skills: The data generated by AMT provides valuable insights into the performance of logistics operations. Consequently, logistics workers need to have analytical skills to interpret this data and make informed decisions based on the insights gained. — Communication skills: Algorithmic management tools are changing the way logistics workers communicate with each other and with their superiors. For example, some logistics companies are using chatbots to talk with their employees and provide them with real-time updates on their work. — Adaptability: The use of AMT requires logistics workers to be adaptable and flexible. They need to be able to quickly learn and adapt to new tools and processes as they are implemented. — Multitasking: Algorithmic management tools are helping logistics workers to become more efficient and productive by automating routine tasks. As a result, workers need to be able to multitask and handle multiple tasks simultaneously to maximize their productivity. 5 Working conditions 5.1 Work intensification Algorithmic management tools can increase work intensity for logistics workers, because workers may be required to complete tasks more quickly or with greater accuracy, as the tool prioritizes 17 speed and efficiency. This can lead to increased stress and burnout among workers. Moreover, algorithmic management tools can disrupt working conditions for logistics workers. For example, workers guided by AI-powered route planning tools may be required to work longer hours or on a more flexible schedule in order to meet the demands of the tool. 5.2 Autonomy The use of AMT can reduce the autonomy of logistics workers, since workers may be given less control over their schedules and tasks, as the algorithm determines the most efficient way to complete the work. This can happen, for example, when workforce direction and task allocation are managed through instructions delivered via handheld or wearable digital devices. Workers use scanners to transmit real-time operational updates, while their devices simultaneously track metrics such as task completion speed and efficiency. This data feeds into central algorithms that analyse variables like location, movement, and timing to streamline workflows and assign tasks accordingly. As a result, workers have minimal flexibility to challenge or modify instructions, as algorithmic decisions are predetermined to optimize efficiency 29 . This can also lead to workers feeling less engaged and less satisfied with their jobs, and, as a consequence, to a loss of flexibility for the organization 30 . 5.3 Social interaction Algorithmic management tools can impact interactions between logistics workers and their colleagues, supervisors, and customers. For instance, workers may be required to communicate more frequently and more quickly in order to meet the demands of the tool. On the other hand, workers may interact less with their colleagues, as they are focused on completing their individual tasks as efficiently as possible. This can lead to a more isolated and impersonal work environment 31 , 32 . 6 Ethics of AI The ethical issues of algorithmic management tools were further discussed with a university professor at Paris-Sorbonne University specialised in AI and ethics. She affirmed that companies using AMTs must invest resources in acquiring experts in quality of work. This is to ensure the recommended methodologies for understanding ethical risks in a multi-agent, human-machine system are met. Once a moral mapping of the situation is made, this allows those who build solutions to do so 'ethically by design' to control a certain number of ethical characteristics and then monitor cooperation between humans and machines. Ultimately, external stakeholders must recommend audits with set criteria, and humans to check the causes of bias, the professor said. We cannot underestimate the risks of AI, she said, we have to accept them, but also be more creative. The EU AI Act has characterised the classification of risks, so we must not minimise the risks nor infringe on the creativity of companies. We are working on the issue of ‘trustworthy AI’ and linguistic nudges. These are mechanisms that we describe to observe ethical tensions. The digital giants used to have ethics committees, but they are in the process of firing them as part of their recent redundancy plans. AI and Ethics Professor, Paris-Sorbonne University 29 Rani, U., Pesole, A. and Gonzalez Vazquez, I., Algorithmic Management practices in regular workplaces: case studies in logistics and healthcare, Publications Office of the European Union, Luxembourg, 2024, doi:10.2760/712475, JRC136063. 30 Friedman, Andrew. 1977. ‘Responsible Autonomy Versus Direct Control Over the Labour Process’. Capital & Class, 1(1): 43–57. https://doi-org/10.1177/030981687700100104. 31 Gaborieau, David. 2012. ‘« Le nez dans le micro ». Répercussions du travail sous commande vocale dans les entrepôts de la grande distribution alimentaire’. La nouvelle revue du travail, no. 1 (December). https://doi.org/10.4000/nrt.240. 32 Massimo, Francesco S. 2020a. ‘Burocrazie Algoritmiche. Limiti e Astuzie Della Razionalizzazione Digitale in Due Stabilimenti Amazon’. Etnografia e Ricerca Qualitativa, no. 1/2020: 53–78. https://doi.org/10.3240/96824. 18 The professor stressed the point that an emergence of a dynamic Europe to address ‘trustworthy AI’ was growing, with increasing research. We need to create fundamental and applied research positions within the industry to rediscover the link with the general public because we see gaps being made, she mentioned. Similarly, following an interview with a French lawyer who co-founded a firm specialised in AI technologies, he explained that the legal world is slowly integrating algorithmic management into their activities for tasks that can be automated. These include administrative tasks, contracts, forecasting and client relations. Until recently we had an old way of doing advocacy and now we are starting to make new moves to make it more technological, and more modern. This allows small law firms like my own to limit low value-added interactions and use machines to increase efficiency instead. Law firm specialized in AI technologies, Lawyer Although the firm’s main customer cases are on data privacy and blockchain, these topics were found to be relevant to the logistics industry. Indeed, there are many legal problems with AI tools and algorithmic management, such as questions on data use and privacy or questions surrounding responsibility and liability. For example, the lawyer described the dilemma of responsibility referring to the human behind a specific algorithmic management tool, the manufacturer of the product, and the designer of the algorithm. The lawyer said there has been a growing concern about the liability of algorithmic management tools across the logistics sector. He said that if a technological problem arises from the use of an AMT that has been outsourced to another company, the responsibility will not always be clear. There are important legal and ethical issues, whether the original company or the outsourcing company is accountable. 7 Surveillance and data privacy issues in the era of algorithmic management AMT may track workers' movements and behaviours, leading to increased surveillance, and, as a consequence, increased pressure, and stress among the workforces. For example, some companies have reported that the use of algorithmic management tools such as handheld and wearable devices has led to increased surveillance of workers, with GPS tracking and real-time monitoring of performance. This can create stress and anxiety for workers and may lead to a more rigid and tightly controlled work environment. Similarly, AI-powered route planning tools can also lead to extensive monitoring and surveillance of drivers. Box 3. Case study on French Logistics 4 A key application of algorithmic management tools in the logistics sector is concentrated in Transport Management System companies (TMS). French Logistics 4 is a TMS software company that sells Software as a Service (SAAS) through paying subscriptions available in 4 languages and 6 countries, with its main activity based in France. French Logistics 4 creates the technology behind a dashboard that other companies use for managing their logistics. Clients include distributors for delivery services, carriers for courier services, or shippers who manage their fleet using the solution, which allows clients to build their logistics platform. The company has developed their own route optimisation engine based on open source. With the rapid rise of delivery services since 2016, our first clients were demanding a web portal to easily manage their logistics, along with a strong technology experience that many other competitors didn’t have. We thought why keep the technology just for us, but instead make it available to other delivery players? This is how the tool was created, a new company, with more tech efforts to serve people for interregional deliveries. French Logistics 4, Founder 19 The tool is purely a logistics management software that allows users to optimise their delivery routes according to the drivers located nearby or other chosen factors. French Logistics 4 customises the platform to the needs and themes of its customers and acts like an Enterprise Resource Planning (ERP) for delivery and logistics companies, or for non-delivery companies wishing to start delivery services. The first step in using the tool is to choose the delivery times, broken down by day and hour, and delineate the zone of activity, which can be a neighbourhood, an entire city, or a region. Afterwards, the dashboard gives many options for users to decide how to build their services. Customers can choose their desired services, area of interest, the tariffs per km/minute/distance, type of vehicle to be used, availability of services (time window frames) etc. Their prices can be set per km, per base, per volume or weight. Figure 3 illustrates the dashboard for a demo company after having chosen all of the above. In this window, users can have an overview of what is going on per delivery, the missions accomplished, those done, dispatched, ongoing, cancelled or failed. Users can also see the delivery missions, time and location of deliveries, status, time taken, number of goods delivered, environmental footprint and many other metrics. French Logistics 4 does not do deliveries but only sells the tool behind the dashboard which allows companies to then plan all these topics. Each company has its dashboard, and the delivery workers have their applications connected to the dashboard. Figure 3: Example of dashboard – list of missions Source: Material provided by French Logistics 4 Figure 4 shows the route tracking window in the tool’s dashboard that allows customers to track each driver, or all at once, and receive live notifications on their status. This helps clients manage their employees and increase internal communication from a centralised area. For example, in an emergency scenario where a hospital must quickly transfer live organs from one point to another, the tool allows one to visualise the status of the trip and its remaining time. This helps trigger the surgical steps earlier and notify the doctor with fewer resources since all information can be received via a phone app. 20 Figure 4: Example of dashboard – route tracking Source: Material provided by French Logistics 4 Figure 5 shows the tool’s performance overview window for all drivers with a subsection for each driver. On top of the window, the customer can visualise the total number of drivers in the company, the drivers mobilised that month, the average turnover per driver that month, and the average turnover per driver the past month. At the centre of the window, the customer can visualise the evolution of new drivers across various months and the top 10 drivers with the most missions completed. At the bottom of the window, the customer can see the list of drivers with each statistic as noted above. Figure 5: Example of dashboard – performance overview of drivers Source: Material provided by French Logistics 4 21 The impacts of using the tool are huge for customers, said the founder. For some retailers, it has allowed them to enter the market and start delivering. Since Covid, many retailers have realised the importance of offering delivery services for their business model, and Covid has boosted customer demand for delivery. For others, it allows them to launch their business faster. The founder noted that their customers can go through them to quickly start their delivery business, which speeds up delivery entrepreneurship. Furthermore, for other companies, it has enabled a dematerialisation process because, before the introduction of tools like this, delivery was based on a paper with information needing a stamp to return to the office for storing and scanning. Finally, for large logistics companies present in several countries, the software has higher stakes for those wanting to have the capacity to manage all of its flows in a single tool. Most companies want to be able to check the progress of operations in a few clicks. The main impact of using the tool is saving time on deliveries and therefore improving the organisation of work and people. The tool saves time because customers no longer have to call the deliverymen to receive updates as they can easily locate them on a map. Less time is wasted on the telephone so unproductive roles are eliminated, which is a financial gain that customers can use elsewhere. In terms of work organisation, it allows customers to gain productivity because it requires fewer people to do more things. Likewise, it is supposed to enhance easier communication between the workers and their supervisors. When asked about possible issues surrounding drivers’ autonomy, as they are constantly being monitored through the GPS application, the founder said that responsibility was for each customer. Normally, delivery drivers have minimum freedom, he said, but after that, we cannot necessarily know what the concrete application is. When asked about relations between business actors and the interactions between players in the sector, the founder said the following. In the fluidity of relations between players in the sector, as the dematerialisation of the sector is slow, distributors are somewhat at the mercy of the transporters’ tools. If I am a company and I have certain distributors with different levels of digitalisation, I cannot offer the same customer experience everywhere I operate. Thus, we allow them to provide their carrier to use our technology where they work. This allows us to work more uniformly. We connect distributors and carriers who use our software. French Logistics 4, Founder Regarding the future of AMTs in the sector, the founder said that these tools will become increasingly essential to keep up with the pace of delivery if companies want to remain competitive. There is a lot of software on the market, some of which is old, he said. The TMS sector is becoming very concentrated with the takeover of start-ups and the use of technologies depends on the way each one operates. Moreover, to gather data on the regulatory challenges of algorithmic management tools, an interview was conducted with the French National Commission on Informatics and Liberty (CNIL). Three experts were interviewed, including a worker at the CNIL’s digital laboratory (LINC), a worker in the solidarity and employment service, which is part of the legal support department, and a sociologist at LINC. The expert working in the legal department said AI tools are particularly expanding in HR and the hiring processes of all companies. Algorithms are increasingly used for application sorting to assist recruiters in analysing audio or video streams and understanding whether candidates are nervous. The subject is very popular and of interest to stakeholders, even those within Logistics. CNIL has published a guide on recruitment applications with certain sheets on the use of algorithmic management. (https://www.cnil.fr/fr/le-guide-du-recrutement: sheet 4 on consent and sheet 13 on algorithmic processing tools). 22 The notion of algorithmic management is of automated decision-making. In Amazon, we see that the principles of the gig economy becoming more and more used for logistics. We can see how the ‘uberization’ of work applies to more and more sectors, including logistics, where historically this was already present. Often, we are asked about surveillance issues at work, which has become possible with widespread video surveillance tools. These are not only used to monitor a place but also people, and since surveillance tools are inexpensive, even SMEs use them with a lack of knowledge of the legal framework. CNIL, AI regulatory expert The issue of worker surveillance remains a central concern, with several high-profile cases drawing public attention. For instance, the CNIL fined SAF Logistics €200,000 for excessive data collection from employees, violating their privacy and failing to cooperate adequately with CNIL services 33 . Similarly, Amazon France Logistique received a €32 million fine for implementing an overly intrusive employee monitoring system, as well as for using video surveillance without proper notification or sufficient security measures 34 . Additionally, during the Covid-19 crisis, the CNIL played a key role in controversies between employers and unions regarding wokplace surveillance. One of the most significant cases involves Amazon France (see Box below). Box 4. A controversy on an AI-based system for social distancing at Amazon France Logistics during the Covid-19 pandemic (2020) During the Covid pandemic, Amazon developed new AI-assisted monitoring tools. One of these was Proxemic a software for visual recognition installed on cameras, whose task is to enhance social distancing by signalling the crowding of workers in the workspace. This was an artificial intelligence system that analysed images from special security cameras and alerted management of potential social distancing violations. Proxemics was built by AI experts in Amazon robotics division and deployed in mid-March in the US. It was introduced progressively in other European countries (1000 Amazon buildings around the world according to Wired). The system consisted of a television screen, depth sensors and an IA-enabled camera, which are installed in different points of the plants. The camera registered images in real-time, tracking people moving through the warehouse. When they pass in the visual field of the camera workers appear on the screen surrounded by “augmented reality” circles. The IA used the apparent size of people in the frame and the number of pixels between them to calculate distance. If the social distance is respected, circles are green; otherwise, circles are red, a possible violation is flagged, and management is alerted. Reviewers include the details in a regular report sent to building managers that summarizes recent social distancing violations in their facility. These features of Proxemics aim, Amazon says, to provide a quick response to contagion risk and are used only for Covid-19 safety. However, this system raised many concerns about privacy on the part of unions and other independent observers, as well as regulation authorities (for instance in France). 33 CNIL. 2023. Excessive data collection and lack of cooperation: the CNIL imposed a sanction on the company SAF LOGISTICS. https://www.cnil.fr/en/excessive-data-collection-and-lack-cooperation-cnil-imposed-sanction-company-saflogistics. 34 CNIL. 2024. Employee monitoring: CNIL fined AMAZON FRANCE LOGISTIQUE €32 million. https://www.cnil.fr/en/employee-monitoring-cnil-fined-amazon-france-logistique-eu32-million. 29 List of boxes Box 1. Case study on French Logistics 1 ............................................................................................................................... 11 Box 2. Case study on French Logistics 3 ............................................................................................................................... 13 Box 3. Case study on French Logistics 4 ............................................................................................................................... 18 Box 4. A controversy on an AI-based system for social distancing at Amazon France Logistics during the Covid-19 pandemic (2020) ......................................................................................................................................... 22 30 Appendix 1. Distribution of interviews Table 1: Distribution of interviews. Interviewee Role/Organisation Date HR and Social Relations Director Anonymous - multinational shipping & receiving and supply chain management company 01/12/2022 Account and Project Manager Anonymous - software and consulting firm 01/12/2022 9 different technology experts and project leaders Anonymous –Independent logistics and supply chain association 02/12/2022 Co-founders and Partners French Logistics 1 09/02/2023 Lawyer, Professor and Tech expert French law firm 09/03/2023 Project officer French logistics foundation 10/03/2023 ‘AI Security’ group leader French logistics association 13/03/2023 CEO French Logistics 3 13/03/2023 Director and Tech Expert French Logistics 2 16/03/2023 Data Scientist Anonymous - multinational container transportation and shipping company 17/03/2023 Computer Science and AI Ethics Professor Paris-Sorbonne University 13/03/2023 Co-founder French Logistics 4 24/03/2023 Technology and AI experts National Commission on Informatics and Liberty (CNIL) 27/03/2023 Source: Authors’ elaboratio Getting in touch with the EU In person All over the European Union there are hundreds of Europe Direct centres. You can find the address of the centre nearest you online (european-union.europa.eu/contact-eu/meet-us_en). On the phone or in writing Europe Direct is a service that answers your questions about the European Union. You can contact this service: — by freephone: 00 800 6 7 8 9 10 11 (certain operators may charge for these calls), — at the following standard number: +32 22999696, — via the following form: european-union.europa.eu/contact-eu/write-us_en. Finding information about the EU Online Information about the European Union in all the official languages of the EU is available on the Europa website (european-union.europa.eu). EU publications You can view or order EU publications at op.europa.eu/en/publications. Multiple copies of free publications can be obtained by contacting Europe Direct or your local documentation centre (european-union.europa.eu/contact-eu/meet-us_en). EU law and related documents For access to legal information from the EU, including all EU law since 1951 in all the official language versions, go to EUR-Lex (eur-lex.europa.eu). EU open data The portal data.europa.eu provides access to open datasets from the EU institutions, bodies and agencies. These can be downloaded and reused for free, for both commercial and noncommercial purposes. The portal also provides access to a wealth of datasets from European countries. The Joint Research Centre (JRC) provides independent, evidence-based knowledge and science, supporting EU policies to positively impact society EU Science Hub Joint-research-centre.ec.europa.eu