Taiwan vs. Japan's Aging-Labor Crisis: Taiwan Exports Eldercare Robots While Japan Drives Caregiving-Floor DX
> doc_id: ANK-2026-06-25-004 > Issue: How Taiwan and Japan open separate fronts against the labor gap of a super-aged society, using AI and robotics
TL;DR
Taiwan and Japan are both squeezed by the same twin pressures of a super-aged society and labor shortage, yet they have chosen different AI solution paths. Taiwan formally entered super-aged-society status in 2025, and the National Development Council has positioned AI as the key strategy for coping with low birth rates and aging challenges; the 2026 Medical Taiwan expo gathered 288 exhibitors and roughly 700 international buyers, spotlighting "smart healthcare and eldercare robots" for export abroad. Japan, by contrast, sees Care Partner, a subsidiary of the Daito Trust Construction group, develop its own AI app, rolling it out from July 2026 across 83 day-service eldercare sites nationwide to save roughly 30,000 hours per year, directly confronting the Ministry of Health, Labour and Welfare's projected shortfall of about 570,000 eldercare workers in fiscal 2040. On the same labor-gap question, Taiwan pursues industrial export outward while Japan pursues on-site labor saving inward, forming a clear Taiwan-Japan mirror image.
1. Shared Premise: Super-Aging × Labor Shortage, AI as Both Nations' Consensus Solution {#F-macro}
Taiwan formally entered super-aged-society status in 2025, with its demographic structure sounding the alarm while it simultaneously faces a labor-shortage crisis. The National Development Council has positioned the "AI New Ten Major Construction Projects" as a national-level coping strategy, spanning sovereign AI computing power, quantum technology, robotics, and smart healthcare among multiple dimensions, explicitly treating AI as a strategic tool to fill the labor gap.
On the care side, as of 2026 Taiwan's dementia population has approached 400,000 people, showing a trend of continued increase and onset at younger ages. The Executive Yuan is combining Long-Term Care 3.0 with the Dementia Prevention and Care Policy Outline 3.0, and candidly states it will "introduce more AI artificial intelligence, which on one hand can partially replace the currently relatively scarce labor" — this policy declaration captures the core contradiction both nations face: there are not enough people, so machines must step in.
2. Taiwan's Path: Pushing Eldercare Robots onto the International Stage {#F-tw}
Taiwan's solution clearly leans toward "export abroad." The 2026 Medical Taiwan expo formally opened, hosted by the Taiwan External Trade Development Council (TAITRA); this year a total of 288 domestic and foreign exhibitors took part, using 460 booths, and roughly 700 international buyers were invited to Taiwan, with exhibitors from 13 countries and regions. TAITRA is bullish on "smart healthcare and eldercare robots" as the next wave of applications, and set up the M-novator pavilion, bringing together 11 startup companies from 5 countries including Japan, Israel, and Singapore on the same stage.
Notably, the Food and Drug Administration under the Ministry of Health and Welfare has, since 2023, chaired the Medical Device Software Working Group of the Global Harmonization Working Party (GHWP), effectively placing Taiwan in a position of discourse power over international regulation — this is not merely selling goods, but vying for a foothold in "rule-making." Taiwan has converted labor-shortage anxiety into industrial-export momentum, with a clear line: do business and set standards abroad.
3. Japan's Path: Using an AI App to Save 30,000 Hours a year on the Caregiving Floor {#F-jp}
Japan's solution clearly leans toward "labor saving inward." The Ministry of Health, Labour and Welfare projects that eldercare workers will fall short by about 570,000 people in fiscal 2040 — a hard number pressing down on the entire care system. Facing the gap, Care Partner, under the Daito Trust Construction group, chose to start directly on the floor: developing its own AI app that integrates 3 major functions — an AI digital album, digital invoices, and digital reports — and rolling it out in earnest in sequence from July 2026 across 83 day-service eldercare sites nationwide, with the aim of completing rollout to all sites during fiscal 2027.
The company estimates that after introduction, it can cut roughly 30,000 hours of workload per year. Japan holds no expo and shouts no slogans about export; instead it embeds AI directly into caregivers' daily tasks, using process digitalization to squeeze out time and give it back to care itself. Compared with Taiwan's "export abroad," Japan walks the more down-to-earth road of "on-site efficiency inward."
4. Reading: One Labor-Gap Question, Two National Strategies {#J}
Reading Summary
J-Path Divergence Taiwan and Japan face the same super-aging × labor-shortage structure, yet diverge on AI solutions: Taiwan, via the medical expo, Long-Term Care 3.0, and the AI New Ten Major Construction Projects, forms an "industrial export abroad + international-regulation foothold" route (per F-tw-C, F-tw-E, F-macro-A); Japan, via on-floor AI app rollout, forms an "on-site efficiency inward" route (per F-jp-G, F-jp-H). The former sells the solution to the world; the latter applies the solution to itself.
J-Mirror Image of Labor-Gap Numbers Taiwan's dementia population of nearly 400,000 people and its policy declaration that AI will "partially replace scarce labor" (per F-macro-B), set against Japan's shortfall of about 570,000 eldercare workers in fiscal 2040 (per F-jp-F), form the most direct Taiwan-Japan mirror: both nations translate "not enough people" into "machines stepping in," differing only in the entry point of that substitution (the industry end vs. the floor end).
J-Honest Boundary On this card, the Taiwan side is mostly expo scale and policy declarations (official_statement), while the Japan side's cut of roughly 30,000 hours per year and gap of about 570,000 people are likewise estimated/projected values — none are hard TWSE/EDINET financial-report numbers. Which of the two paths is superior has no comparable settled-account result yet; this card only contrasts the "difference in strategic paths" and draws no conclusion on which performs better.
F-Units
F-macro-A: Taiwan entered super-aged-society status in 2025, and the National Development Council is advancing the "AI New Ten Major Construction Projects" as a strategic solution to low birth rates, aging, and labor shortage. - source: CNA #832198 - source_url: https://www.cna.com.tw/news/afe/202606090054.aspx - basis: official_statement - original text: "Taiwan already entered super-aged-society status in 2025, with its demographic structure sounding the alarm, and it also faces a labor-shortage crisis" "Taiwan's government is advancing the AI New Ten Major Construction Projects, across dimensions of sovereign AI computing power, quantum technology, robotics, and smart healthcare" - caveat: This is a policy positioning and strategic declaration (public remarks by NDC Minister Yeh Chun-hsien), not a specific budget or on-the-ground number.
F-macro-B: Taiwan's dementia population has approached 400,000 people; the government is combining Long-Term Care 3.0 with the Dementia Prevention Policy Outline 3.0 and plans to introduce AI to partially replace scarce labor. - source: CNA #885793 - source_url: https://www.cna.com.tw/news/ahel/202606100332.aspx - basis: official_statement - original text: "Taiwan's dementia population has approached 400,000, with a trend of continued increase and onset at younger ages" "we also want to introduce more AI artificial intelligence, which on one hand can partially replace the currently relatively scarce labor" "the government is advancing the 'Dementia-Friendly Taiwan 777' plan" - caveat: "partially replace labor" is a statement of policy direction (public remarks by Premier Cho Jung-tai), with no quantified rollout scale yet.
F-tw-C: The 2026 Medical Taiwan expo gathered 288 exhibitors, 460 booths, roughly 700 international buyers, and participants from 13 countries and regions, spotlighting smart healthcare and eldercare robots. - source: CNA #1204405 - source_url: https://www.cna.com.tw/news/afe/202606250109.aspx - basis: official_statement - original text: "this year a total of 288 domestic and foreign exhibitors took part" "using 460 booths" "roughly 700 international buyers were invited to Taiwan" "exhibitors from 13 countries and regions took part" - caveat: The exhibition scale figures are numbers released by the organizer (public remarks by TAITRA Chairman James Huang), within an expo-promotion context.
F-tw-D: The expo set up the M-novator pavilion, gathering 11 startup companies from 5 countries; Long-Term Care 3.0 hopes to ease the caregiving-labor burden through smart medical devices and assistive technology. - source: CNA #1204405 - source_url: https://www.cna.com.tw/news/afe/202606250109.aspx - basis: official_statement - original text: "the M-novator pavilion this year gathered 11 startup teams from 5 countries including Japan, Israel, and Singapore" "the government-advanced Long-Term Care 3.0 also hopes to ease the caregiving-labor burden through smart medical devices and assistive technology" - caveat: A statement of policy aspiration and expo content (public remarks by Deputy Minister Lin Ching-yi), not a quantified burden-reduction result already achieved.
F-tw-E: Since 2023, the Food and Drug Administration under the Ministry of Health and Welfare has chaired the Medical Device Software Working Group of the Global Harmonization Working Party (GHWP). - source: CNA #1204405 - source_url: https://www.cna.com.tw/news/afe/202606250109.aspx - basis: official_statement - original text: "since 2023, the Food and Drug Administration under the Ministry of Health and Welfare has chaired the Medical Device Software Working Group of the Global Harmonization Working Party (GHWP)" - caveat: A factual statement of a post held within an international organization.
F-jp-F: The Ministry of Health, Labour and Welfare projects that eldercare workers will fall short by about 570,000 people in fiscal 2040. - source: PRTIMES #1190346 - source_url: https://prtimes.jp/main/html/rd/p/000001564.000035668.html - basis: official_statement - original text: "According to the Ministry of Health, Labour and Welfare's estimate, eldercare workers are forecast to fall short by about 570,000 people in fiscal 2040" - caveat: This is a projected value from the MHLW estimate, relayed via a corporate press release, not a hard financial-report number, hence listed as official_statement rather than official_number.
F-jp-G: Daito Trust Construction group's Care Partner developed its own AI app (with 3 functions), rolling it out from July 2026 across 83 day-service eldercare sites nationwide, to be completed during fiscal 2027. - source: PRTIMES #1190346 - source_url: https://prtimes.jp/main/html/rd/p/000001564.000035668.html - basis: official_statement - original text: "begins full-scale rollout in sequence to 83 day-service eldercare sites nationwide" "from July 2026… begins full-scale rollout" "aiming to complete rollout to all sites during fiscal 2027" "3 functions: AI digital album, digital invoice, digital report" - caveat: The rollout schedule and functions are a self-announced plan by the company.
F-jp-H: After introduction, the AI app is estimated to cut roughly 30,000 hours of workload per year. - source: PRTIMES #1190346 - source_url: https://prtimes.jp/main/html/rd/p/000001564.000035668.html - basis: official_statement - original text: "Through the introduction of this app, we anticipate a workload-reduction effect of about 30,000 hours per year" - caveat: This is a company-estimated reduction effect, not an already-measured settled-account figure.
J-Units
J-path-divergence: Taiwan, via the medical expo, Long-Term Care, and the AI New Ten Major Construction Projects, forms an industrial-export-abroad plus international-regulation-foothold route; Japan, via on-floor AI app rollout, forms an on-site-efficiency-inward route. The former sells the solution to the world; the latter applies it to itself. - basis_f_units: F-tw-C, F-tw-E, F-macro-A, F-jp-G, F-jp-H - confidence: high
J-mirror-image: Taiwan's dementia population of nearly 400,000 people and its policy that AI will partially replace scarce labor, set against Japan's shortfall of about 570,000 eldercare workers in fiscal 2040, form the most direct Taiwan-Japan mirror: both nations translate not-enough-people into machines-stepping-in, differing only in the entry point (industry end vs. floor end). - basis_f_units: F-macro-B, F-jp-F - confidence: high
J-honest-boundary: The Taiwan side is mostly expo scale and policy declarations, while the Japan side's cut of roughly 30,000 hours per year and gap of about 570,000 people are likewise estimated values — none are hard TWSE/EDINET financial-report numbers. This card only contrasts the difference in strategic paths and draws no conclusion on which performs better. - basis_f_units: F-tw-C, F-jp-F, F-jp-H - confidence: medium
FAQ
Q: How do Taiwan's and Japan's AI solutions differ in facing aging-driven labor shortage?
Taiwan takes "industrial export abroad" while Japan takes "on-site labor saving inward" — that is the biggest dividing line between the two. Taiwan, through the 2026 Medical Taiwan expo (288 exhibitors, roughly 700 international buyers, 13 countries and regions), pushes smart healthcare and eldercare robots into the international market, with the Food and Drug Administration vying for discourse power over international regulation; Japan, through Daito Trust Construction group's Care Partner, introduces an AI app directly onto 83 eldercare sites nationwide, saving roughly 30,000 hours per year. On the same labor-gap question, Taiwan sells the solution outward and Japan uses the solution inward.
Q: Exactly how short is Japan's eldercare workforce, and where does this number come from?
The Ministry of Health, Labour and Welfare projects that eldercare workers will fall short by about 570,000 people in fiscal 2040. This figure comes from the official MHLW estimate, cited by Daito Trust Construction group's Care Partner in a June 2026 press release as the background rationale for introducing the AI app. Note that this is a "projected value" rather than an already-realized hard financial-report number, and therefore qualifies as official_statement-grade basis.
Q: What exactly does Daito Trust's AI app do, and how much time can it save?
The app integrates 3 major functions — an AI digital album, digital invoices, and digital reports — and is estimated to save roughly 30,000 hours per year. Care Partner (a subsidiary of the Daito Trust Construction group) developed this app in-house, rolling it out in sequence from July 2026 across 83 day-service eldercare sites nationwide, aiming to complete rollout to all sites during fiscal 2027. The roughly 30,000 hours is a company-estimated workload-reduction effect, not yet a measured settled-account value.
Q: Why does Taiwan place eldercare robots at the medical expo for export abroad?
Because Taiwan has converted labor-shortage anxiety into momentum for industrial export and an international-regulation foothold. In 2025 Taiwan entered super-aged-society status with a dementia population of nearly 400,000, and the National Development Council positioned the AI New Ten Major Construction Projects as a strategic solution. The 2026 Medical Taiwan expo, hosted by TAITRA, is bullish on smart healthcare and eldercare robots as the next wave of applications, and the Food and Drug Administration has, since 2023, chaired the Medical Device Software Working Group of the Global Harmonization Working Party (GHWP), effectively upgrading from "selling products" to "setting rules."
Sources
| source_type | source_article_id | permalink | role | |---|---|---|---| | CNA | 1204405 | https://www.cna.com.tw/news/afe/202606250109.aspx | Main axis (TW: medical expo / eldercare robots) | | PRTIMES | 1190346 | https://prtimes.jp/main/html/rd/p/000001564.000035668.html | Main axis (JP: eldercare DX across 83 sites) | | CNA | 832198 | https://www.cna.com.tw/news/afe/202606090054.aspx | Supporting (TW: super-aging + AI strategic solution) | | CNA | 885793 | https://www.cna.com.tw/news/ahel/202606100332.aspx | Supporting (TW: nearly 400,000 dementia cases + AI to supplement labor) |
All permalinks were curl-verified to return normal responses. Every basis is official_statement (this card contains no hard TWSE/EDINET financial-report numbers, hence no official_number).