1. What was actually measured?

Intel published the findings on August 20, 2026. The independent opinion research was commissioned by Intel, designed with Man Bites Dog and fielded by Coleman Parkes in 2026. Its 800 respondents worked at organizations with at least $500 million in annual revenue across the US, China, Japan, Germany, the UK and South Korea.

This is not a random survey of all companies, excludes the Middle East, and includes 430 robotics specialists. It describes the expectations of large-organization buyers and practitioners in six markets; it does not prove that 60% of companies worldwide will run fleets within five years.

  • Measured: stated expectations, confidence and current plans.
  • Not measured: purchased units or audited return on investment.
  • Sponsor: Intel, a robotics-compute supplier.
  • Best use: an early indicator of enterprise intent.

2. Intent is not deployment: put IFR beside the survey

Six in ten respondents expect a fleet within five years, while leaders on average predict full-scale deployment could double output. “Could” matters: the report presents an expected scenario, not a controlled trial or audited financial result.

IFR records a different thing—completed market activity. It counted 542,000 industrial robots installed during 2024 and 4.664 million operating worldwide at year-end; Asia took 74% of new installations. Those are market statistics, but industrial robots are narrower than the report's retail, healthcare and smart-city scope.

3. The strategy gap is an ownership and skills gap

Although 67% believe they will be ready for a mixed workforce by 2030, just 40% have a formal strategy today. Primary ownership sits with the CTO at 29%, CIO at 18%, COO at 16%, and a dedicated robotics or automation leader at only 12%. Fragmented accountability explains how a demo can work while a fleet stalls.

Three quarters say new human roles will be required, 41% say HR leaders are not equipped to plan for a robot-inclusive workforce, and four in ten cite skills as a scaling blocker. Edge and on-device AI, maintenance, fleet operations, data and cybersecurity all need operating capability—not just vendor access.

  • Name an owner for the business outcome, not only the machine.
  • Design operation, maintenance and incident roles before purchase.
  • Tie training to shifts and measurable performance.
  • Contract for knowledge transfer, not permanent dependency.

4. Safety is not a certificate on the arm

Fifty-five percent say safety concerns delayed deployment, while 48% say their organization experienced a safety incident involving a deployed robot. The latter is self-reported and lacks severity, definition and period, so it is not an official injury rate. It still signals that system safety cannot be reduced to product certification.

ISO 10218-1:2025 addresses the robot as machinery; ISO 10218-2:2025 covers integration, commissioning, operation, maintenance and decommissioning of the application or cell. NIOSH adds the worker context for collaborative, mobile, wearable and future AI-enabled robots. Acceptance testing must combine path, tool, payload, software, operator and failure states.

5. The Saudi factory translation

Saudi Arabia's Future Factories program links transformation to maturity assessment, advanced technology and operating efficiency—not a single robot purchase. Start with one constrained bottleneck such as quality inspection, hazardous handling or internal transport, then compare human operation, conventional automation and an AI-enabled robotic solution.

Test heat, dust, shifts and real network conditions; decide what runs on-device and what can depend on cloud services. The survey says 42% have a funded edge/on-device strategy and 39% are developing one, but these are stated plans. The decisive evidence is a documented safe-stop and local-decision test when connectivity fails.

  • Baseline units, defects, downtime and incidents per shift.
  • Count integration, training, maintenance, spares, energy and software.
  • Test the application and cell, including stop and recovery drills.
  • Define roles, skills and escalation before expanding to ten machines.

6. The gate from one robot to a fleet

Do not begin with form factor. Seventy-seven percent of respondents prioritize performance over appearance, and a fixed arm, mobile platform or inspection system may deliver a clearer return than a general humanoid. Define the task, environment, accuracy, latency and tolerated downtime first.

Expansion should require four proofs: sustained improvement against baseline, no unacceptable residual-risk increase, an internal team able to operate and maintain every shift, and a contract covering spares, updates, security and exit. Only then does a fleet become an operating decision instead of repeated procurement of a successful demo.