Polish industry is entering the second half of 2026 with stronger output, but without the comfort of a stable economic climate. In June, output for sale rose by 7.6 per cent year-on-year (GUS), but after seasonal adjustment, it increased by only 0.1 per cent compared with May. Across the European Union as a whole, industrial production in May was 0.3 per cent lower than a year earlier (Eurostat). Companies today need modernisation measures that rapidly improve costs, quality or throughput. Long-term programmes with deferred returns are becoming harder to justify.
Cost pressures are also continuing. In the first quarter of 2026, hourly labour costs in EU industry rose by 3.6 per cent. At the same time, Europe installed 85,000 industrial robots in 2024, 8 per cent fewer than the previous year, although this was the second-best result on record (IFR). Capital has not disappeared, but has become more selective. Projects must justify themselves through short implementation times, limited downtime risk and the ability to scale up to further production lines.
Brownfield rather than a showcase factory
Modernisation most often concerns plants that have been in operation for a dozen or so years, or even several decades. Machines from different generations, controllers from multiple manufacturers and applications whose documentation no longer corresponds to the actual configuration all operate within a single production hall. Under such conditions, automation without halting production means shifting the design, testing and the majority of errors away from the production line.
Work begins with collecting data and identifying the actual process constraint. This could be a bottleneck, changeover time, defect rate, micro-downtime or the failure rate of a specific piece of equipment. Without this diagnosis, automation merely accelerates a poorly designed process.
New control logic, robot programmes and operator interfaces are then tested in a simulation or on a digital twin. The system is first deployed on a section of the line, operates in parallel with the old configuration and is switched over during a short, scheduled window. The project must have a rollback scenario in place.
Wipro PARI applied this model during the expansion of an operational engine assembly line. The virtual commissioning reduced on-site work by 70 per cent, cut the need for rectifications by 40–50 per cent, and the entire project was completed in three months instead of the usual six to eight. The tests covered PLC code, robots, HMIs, safety interlocks and 17 product variants. This is a study by the technology supplier, not an independent audit. However, it clearly illustrates the source of the savings: errors are rectified before the integrators enter the production hall, rather than during commissioning carried out under production pressure.
The result comes from the process, not from a robot catalogue
The robot itself does not eliminate process variability, inconsistent data or organisational errors. Jubilant Ingrevia modernised an operational chemical plant by implementing over 30 interlinked applications involving AI, machine learning, the Industrial Internet of Things, digital twins and predictive analytics. According to the Global Lighthouse Network, process variability fell by 60 per cent, whilst production volume almost doubled. This result was achieved through the integration of multiple use cases, rather than the installation of a single device.
A similar approach was adopted by Rockwell Automation’s facility in Singapore, which handles over a thousand SKUs and carries out more than 20,000 changeovers annually. Over 50 digital and AI solutions, covering flexible automation, quality control and smart maintenance, increased the number of units per man-hour by 43 per cent. The number of defects fell by 35 per cent, whilst the time taken for staff to become self-sufficient was reduced by 67 per cent.
The results of plants belonging to the WEF network are not a universal benchmark. They relate to mature programmes implemented by companies with the necessary capital, expertise and scale. However, they demonstrate a consistent pattern: the highest return is achieved when the technology simultaneously transforms the management of flow, quality and maintenance.
The CIO is responsible for production continuity
In a brownfield project, the CIO’s role extends beyond the provision of infrastructure and data. Configuration versioning, copies of control programmes, OT network segmentation, access control and an up-to-date map of dependencies between devices and systems are all required. Without these, every change increases the risk of downtime, and the digital model quickly ceases to correspond to the physical production line.
The board should evaluate the project in terms of the value of production recovered: the number of minutes of downtime, changeover time, scrap rates, energy consumption per unit, and the margin lost due to bottlenecks. The number of robots, sensors or AI models reflects the cost and scope of the investment, not its outcome.
Poland remains a market where businesses have limited digital maturity. In 2025, 8.4 per cent of companies were using AI technologies, compared with 20 per cent across the EU as a whole. A 2026 report by the European Commission continues to point to the slow uptake of advanced technologies and stagnation in the number of ICT specialists. After the summer, the pace of automation will therefore depend primarily on the ability to integrate, test and maintain changes within existing production processes.
