KEF survey on physical AI adoption and awareness
47% currently using it, 37% planning to adopt
62% of respondents say physical AI will reduce overall jobs
More than 8 in 10 large South Korean manufacturers have already adopted or plan to adopt "physical AI" — a category that includes robots and self-driving cars — and companies expect the technology to boost productivity by an average of 31.9 percent.
The Korea Enterprises Federation said Thursday it had surveyed 100 manufacturers with 500 or more employees on the adoption and awareness of physical AI.
Physical AI refers to technology that combines artificial intelligence with physical systems such as robots and automated equipment to perform real-world tasks. Representative examples include humanoid robots, self-driving cars, drones, automated guided vehicles (AGVs) and autonomous mobile robots (AMRs).
The survey found that 84 percent of respondents had already introduced physical AI or planned to do so. Of those currently using it, 4 percent had deployed it across all processes or departments and 43 percent in select processes or departments. An additional 37 percent had not yet adopted physical AI but planned to, while only 16 percent said they had no current or future plans to do so.
Larger companies showed a stronger inclination to adopt the technology. Among firms with 1,000 or more employees, 88.5 percent had adopted or planned to adopt physical AI, compared with 68.2 percent among those with between 500 and fewer than 1,000 employees.
Among companies currently using physical AI, the AGV/AMR type was the most widely deployed at 63.8 percent, followed by self-driving vehicles at 31.9 percent, humanoid robots at 29.8 percent and drones at 4.3 percent.
Among companies planning future adoption, however, humanoid robots topped the list at 64.9 percent, followed by self-driving vehicles at 37.8 percent, AGV/AMR types at 35.1 percent and drones at 8.1 percent. The Korea Enterprises Federation said the results suggest that demand for humanoid robots is likely to grow across a wide range of sectors.
Assembly and manufacturing processes accounted for the largest share of current physical AI applications at 66 percent, with logistics, transportation and warehouse management close behind at 57.4 percent. Other areas included product design and development at 17 percent, quality diagnostics at 14.9 percent and safety management at 10.6 percent, indicating that the technology's reach is broadening.
When the 84 companies that had adopted or planned to adopt physical AI were asked about expected benefits, 45.2 percent cited productivity and earnings improvement as the top anticipated effect, followed by relief from labor shortages and workforce management difficulties at 26.2 percent and prevention of industrial accidents and reduction of safety risks at 22.6 percent. These companies expected physical AI to improve productivity by an average of 31.9 percent.
Concerns about job losses were widespread, however. Some 61.9 percent of respondents said the job-displacement effect of physical AI adoption and expansion would outpace new job creation, resulting in an overall reduction in employment. Another 34.5 percent said there would be little change, while 3.6 percent expected employment to grow.
The biggest challenge anticipated in adopting and using physical AI was technological instability and system-failure risks, cited by 34.5 percent of respondents. Uncertainty over return on investment came in at 31 percent, while opposition from unions and workers stood at 27.4 percent.
Despite these concerns, companies broadly expected physical AI to play a significant role in addressing national economic challenges such as a shrinking workforce and declining productivity. Some 79.8 percent said the technology's spread would have a large positive impact on resolving such problems, while 15.5 percent said the impact would be moderate and 4.8 percent said it would be small.
"Physical AI is expected to spread further, and its impact on companies and the national economy is likely to grow significantly," said Lee Seung-yong, head of the Korea Enterprises Federation's economic analysis team. "Concerns are also being raised about uncertainty in returns on investment and potential labor-management conflict." He added that relevant regulations should be carefully reviewed and active corporate support measures put in place to maximize the positive effects physical AI can bring.
eyre@heraldcorp.com
