In 2026, France Travail counts 3,040 non-seasonal recruitment projects for electrical, electronic and automation maintenance technicians; this statistic does not measure vacant positions. The French robotics market is worth 4.5 billion euros and continues to grow, but companies that want to take the leap encounter recruitment difficulties for machine installation, programming and operation roles. Training and coordination of stakeholders represent among other issues, without proof that they alone explain the bottleneck.
The essentials
- France Travail documents recruitment projects under tension in maintenance and automation sectors.
- Robotics density follows an unequal territorial distribution.
- Germany and South Korea have developed distinct approaches: decentralized dual training for one, concentrated training clusters for the other.
- According to Acemoglu, access to training in robotics and automation matters for technology deployment.
- Steering training policy toward decentralization constitutes a structural issue for access to skills.
5,000 vacant positions in a booming market
The figure strikes you with its irony. France has a robotics industry that is accelerating. The order books of integrators are overflowing. Manufacturing SMEs, under competitive pressure, are seeking to automate. And yet, thousands of recruitments are projected in certain maintenance and automation families, with high recruitment difficulties.
A robotics integrator is the link between the machine and the production line. They program articulated arms, configure vision systems, diagnose breakdowns and train operators. It’s a hybrid profession, at the intersection of electrotechnics, industrial computing and mechanics. Vocational high schools produce few of them. BTSs (technician diplomas) and professional licenses struggle to calibrate their intake against actual demand.
France Travail records 3,040 non-seasonal projects in 2026 for a family of maintenance and automation technicians, with no established estimate of graduates.
Market growth compounds the equation. According to current trends, the stock of installed robots is experiencing steady expansion. Growth in the fleet could increase maintenance and programming needs, but their increase is neither automatic nor independent of new installations. A company that automated its line five years ago must now upgrade it, and is seeking skills that the market does not supply.
Dares documents durably high tensions in maintenance professions; precise attribution to France Stratégie and solely to automation professions has not been established.
The map of robots follows the map of training
The national deficit is accompanied by marked territorial disparities. French industrial robotics historically concentrated in automotive basins: the Grand-Est around Mulhouse and Strasbourg, Bourgogne-Franche-Comté around Belfort and Montbéliard, the Pays de la Loire with Renault and Stellantis subcontractors. These regions have ecosystems of training and industrial subcontracting.
The rest of the territory follows an inverse logic. According to several sector observers, SMEs located outside historical industrial basins encounter obstacles in accessing robotics integrators. They may reach a Paris or Lyon-based firm, but costs explode and technical follow-up becomes problematic. These companies encounter barriers to automation.
The IFR documents national and sectoral robotics densities, without systematic data by French region. In 2023, France has 186 robots per 10,000 industrial employees according to the IFR, trailing Germany (429 in 2023) and South Korea (higher density). But within the territory, the gap between historical industrial basins and less industrialized regions is probably of the same order of magnitude, without precise sub-regional data being systematically published.
This mechanism connects to a broader dynamic I described in a previous article on the capital-labor divide: productivity gains concentrate where skills and capital are already present. Robotics seems to amplify this movement at the territorial level.
Germany and South Korea facing the same challenges
International comparisons are not flattering, but they are useful because they show that the French shortage is not a natural inevitability. The talent shortage is documented, but causal attribution to institutional and training choices has not been demonstrated by this source.
In 2023, Germany displays a robotics density roughly 2.3 times higher than France’s. Its density of trainers is proportionally higher. The dual system combines apprentices in companies and vocational schools, an important characteristic of German training. Regional Chambers of Commerce certify programs, large companies co-finance teaching equipment. Volkswagen, Siemens and Bosch directly train their future technicians in centers they co-construct with the Länder.
In Germany, the dual system distributes training across the territory.
South Korea chose a different architecture. With 1,000 robots per 10,000 employees, the highest rate in the world—it concentrates training in industrial clusters around Ulsan, Changwon and Gumi. These clusters are closed systems: training, production and R&D within the same geographic perimeter, jointly financed by the state and chaebols. The model is efficient but difficult to replicate outside these zones, creating its own form of territorial concentration. I detailed how this model compares to other Asian approaches in an article on the two industrial robotics models in Asia.
These two paths are different, but they have one point in common: the state or regional institutional players explicitly decided to organize training based on deployment needs, rather than letting educational supply spontaneously adjust to the market.
France has attempted isolated initiatives: the Campuses of Trades and Qualifications labeled by the Education Ministry include a few centers on automation. Some Campuses of Trades and Qualifications do cover mechatronics, robotics and automation; their insufficiency of resources, territorial articulation or visibility has not been established by a consulted evaluation.
The institutional bottleneck according to Acemoglu
Economist Daron Acemoglu contends that AI complementary to workers, capable of augmenting their capabilities rather than simply replacing tasks, remains under-invested and could orient productivity gains and their distribution more favorably. Acemoglu forecasts potentially modest macroeconomic gains with the observed trajectory and emphasizes the importance of technologies creating new tasks; the institutional causality claimed is not demonstrated in this study.
Applied to the French case, this framework is striking. The shortage of robotics integrators creates a situation close to what Acemoglu analyzes: according to this logic, robots would install preferentially in regions where skills and infrastructure already exist, which could limit access to automation in less equipped territories. SMEs that cannot find technicians locally encounter barriers to automation.
A competing reading deserves consideration. Economists close to the liberal supply-side current, such as Philippe Aghion or Tyler Cowen, would be tempted to respond that a skills shortage mechanically generates higher wages for integrators, which attracts more candidates to these professions and stimulates private training supply. The market adjusts, with a lag.
The wages of robotics integrators have indeed risen in France: a confirmed technician can expect 45,000 to 55,000 euros annually, which is attractive to a BTS graduate. Private training centers have opened short courses on collaborative robot programming.
But spontaneous adjustment has its structural limits. First, the lag: training a capable integrator takes between one and three years depending on the level targeted. The training delay represents an obstacle to SME access to automation. Second, geography: the supply of private training varies by territory. An SME in a rural area does not attract a private center.
Finally, the barrier to entry: training as an integrator at a private center costs between 5,000 and 15,000 euros, which neither a career-changer nor a struggling SME can easily finance without public support.
The market corrects at the top. It leaves the bottom waiting.
France’s choices through 2030
The current deficit is not fixed. The question is who will fill it, how, and for whose territories.
The supply of training in automation varies by region, with a more established presence in industrialized areas. Without targeted intervention, territorial productivity disparities risk widening.
A second scenario passes through explicit public redirection, on the model of what France Compétences and the regions attempted in other tense sectors. This would require sizing the Campuses of Trades and Qualifications for automation based on territorial needs, co-financing teaching equipment in vocational training centers outside major urban areas, and articulating vocational training schemes with local SME needs. South Korea’s example shows that targeted investment in skills can transform robotics penetration rates in a decade, but it demands explicit public will and sustainable funding, not annual calls for projects.
A third path, more hybrid, is sketched in a few regional experiments. Auvergne-Rhône-Alpes has developed short-course paths in collaborative robotics with local industrialists. Nouvelle-Aquitaine funds “robotics ambassadors” tasked with guiding SMEs in their first deployment. These initiatives remain modest relative to the scale of the deficit, but they test a model: train quickly, train nearby, train with the companies that will hire. It’s an answer to both the delay problem and the geographic problem at once.
The central question this architecture poses is intergenerational. Unequal access to automation training today risks limiting future job prospects. The competition that South Korea built around its technical vocational high schools shows that these choices are made early, or not at all. Unequal access to teaching equipment in automation creates qualification disparities. And the best-equipped vocational schools are, unsurprisingly, in regions that already were.
Available measures for SMEs
Waiting for large-scale public policy does not suspend business needs. A few levers exist today.
Employer consortiums, a formula allowing several SMEs to share one employee, have been used successfully in food processing and industrial maintenance. Several regions are experimenting with shared integrator consortiums, where a technician accompanies three or four companies from the same basin in rotation. The model divides the cost, reduces wait times and allows SMEs that could never have recruited alone to access the skill.
Robot manufacturers have also evolved their offerings. FANUC, ABB and Universal Robots now offer certifying training courses of two to five days on their equipment, accessible to existing operators and maintenance technicians. These trainings do not cover the full profession of integrator, but they enable rapid skill advancement, internalization of routine operations and reduced dependence on external providers.
These solutions complement public initiatives without constituting a systemic-scale response. They show that private actors and local governments are tackling the problem head-on, and that the choice to rely on them or to build a more ambitious public architecture is precisely the type of institutional arbitrage that Acemoglu places at the center of the debate on the direction of technological progress.
France has an industrial fabric that wants to automate, a dynamic robotics market and an education system that knows how to train technicians. The bottleneck manifests in a failing articulation between the industrial fabric, the robotics market and the French education system. Filling this bottleneck before the gap between territories becomes irreversible: that is the wager that the coming years will or will not allow to hold.
Sources
- France Travail, data on occupations in tension, March 2026, https://www.i-actu.fr/blog/robotique-automatisation-industrie-france/
- International Federation of Robotics (IFR), World Robotics 2025, https://ifr.org/ifr-press-releases/news/robot-race-the-worlds-top-10-automated-countries
- Daron Acemoglu, The Simple Macroeconomics of AI, Economic Policy, 2024, https://doi.org/10.1093/epolic/eiae042
- France Stratégie, employment projections and occupations in tension, reports available at france-strategie.gouv.fr
- Planète Robots, sectoral data on the French robotics market 2024
