Will robots take palletiser jobs?

Threat timeline: 0-2 years across Chinese welding, photovoltaic, lithium-battery, metal fabrication and automotive-parts lines, where displacement is already happening rather than approaching. The reason is not that this machine is advanced — a 20 kg, 1.75 m six-axis arm is a specification so mature it is boring — but that the cost lines have crossed. Guosen Securities models a 10,000-tonne-per-year steel structure line at roughly RMB 3.75M/year for manual welding versus RMB 3.30M/year for robot welding: robots are already RMB 450,000/year cheaper, and the gap is widening. One welding robot displaces two to three welders, and every three robots need one general operator at around RMB 100,000/year, against an average welder salary above RMB 180,000. Estun's roughly 25% shipment growth in 2025 and its 10,000-unit quarter in Q2 2026 say this substitution is moving at the pace of shipments, not press releases. 2-5 years for export markets: 15,000 units/year of Polish capacity plus Carl Cloos's German channel push the same cost curve into European metal fabrication and automotive supply chains. One honest counterweight: Chinese manufacturing has a severe labour shortfall of its own — a joint MIIT/MOHRSS/MOE plan projected a gap of roughly 30 million workers in key manufacturing sectors by 2025, a 48% shortfall rate — so a substantial share of this substitution is filling vacancies rather than pushing incumbents out. For an individual welder the difference is narrower than it sounds: not a layoff, but a hard ceiling placed on the bargaining power and wage growth of the trade by a machine that costs RMB 450,000 less per year.

Threat timeline: 1-3 years, and this machine's displacement mechanism differs enough from the large industrial arms elsewhere in this catalogue to deserve separate treatment. The UR12e weighs 33.5 kg, has a 190 mm base footprint, an IP54 rating, and 17 configurable safety functions built in. Together those specs mean one thing: it can be placed beside an existing workbench without a safety fence, without a dedicated robot engineer, and without rebuilding the line. Deploying a conventional six-axis robot is an engineering project. Deploying a cobot is closer to a purchase order. So the exposure is not the automotive OEM — it is the small and mid-sized manufacturer that could never previously afford automation at all, which is also the layer where employment density is highest. Teradyne Robotics posting a fifth consecutive quarter of growth, its first USD 100M quarter, and US revenue rising to 32% of the segment is evidence that this threshold is actually being crossed. The 3-7 year variable is software, not hardware: Universal Robots maintains both PolyScope 5 and PolyScope X and has been pushing AI-assisted programming through 2026. The dominant cost in a cobot deployment has never been the arm; it is programming and integration. Only once that compresses does the USD 44,636 figure start comparing directly against the annual fully-loaded cost of a workstation. One honest brake on all of this: the newest published cumulative-sales figure is still the 100,000 cobots announced in February 2025, eighteen months without an update — which suggests the category may be expanding rather more slowly than the revenue growth rate implies.