CIEA · Compass Global Magazine · Proposal for Taichung EDB

A Feasibility Study for an Industrial AI Hub in Taichung

A research commission, not a hub funding request. In 2026 the Bureau hires Compass Global Magazine and CIEA, under two separate contracts, to determine whether Taichung has the demand, the manufacturer commitment, and the institutional fit to anchor an international industrial AI hub. The Bureau owns the data, under any outcome.

01The Reframe

The earlier discussion positioned the Bureau as a partner in launching a hub. We are revising the ask. In 2026 we are not asking the Bureau to commit to a hub. We are asking the Bureau to be the first commissioning client of a structured feasibility study that answers, with evidence, whether the hub thesis holds in Taichung. CIEA earns the Bureau's trust by delivering a research product, not by promising an institution. The 2027 conversation about a hub only happens if the data justifies it.

Engagement: 2026 Industrial AI Hub Feasibility Study, commissioned by the Taichung EDB
Vehicle: Procurement (政府採購 / 委託研究), not grant (政府補助)
Term: August 2026 fieldwork begins · Q1 2027 final report delivered
Total budget: NT$1,000,000 – 1,500,000, split across two separate contracts (see §04)
Client ownership: EDB owns the dataset and the final report. CIEA and the Magazine retain only the methodology.

02What the Study Answers

Question 01 · Place

Is Taichung the right city for an international industrial AI hub?

Industrial-density mapping, university capacity audit, infrastructure read, comparison with Taipei / Hsinchu / Tainan as alternative anchors.

Question 02 · Demand

Is there enough manufacturer commitment to sustain the hub past launch?

Founding-sponsor expressions of interest, cert co-design willingness, project-challenge intake, worker-upskilling registration depth.

Question 03 · Shape

If feasible, what does the hub need to look like to actually work?

Partnership architecture, governance model, facility profile, faculty pipeline, sequencing of AWS / NVIDIA / AMD academy applications.

03Methodology — Two Research Layers

Layer A · Oct – Nov 2026

Digital demand instrumentation

Public landing pages for worker upskilling and for founder interest. Structured manufacturer questionnaires routed through Compass Global Magazine's editorial channel. Geographic, role, and segment data captured against a defined registration funnel.

Layer B · December 2026

In-person convening & commitment-testing

Tech-Manufacturer Hub Dinner — 60 to 80 manufacturer CEOs alongside AWS, NVIDIA, AMD, university representatives, and the press. Structured pre- and post-event interviews capture commitment depth, not just attendance.

04Engagement Structure — Two Contracts, Two Scopes

Sole client

Taichung Economic Development Bureau

Contract A · Research

Compass Global Magazine

Designs and executes the intelligence work: market scan, digital instrumentation, manufacturer questionnaires, written report, comparative analysis with peer hubs (AI Singapore, ATDC, MBZUAI).

Indicative · NT$700,000 – 1,000,000

Contract B · Convening

CIEA (社團法人康百視文教交流協會)

Runs the in-person research layer: December dinner, manufacturer outreach and stakeholder interviews, bilingual documentation, post-event commitment readout. Matches CIEA's existing intercultural-exchange charter.

Indicative · NT$300,000 – 500,000

GovernanceEach contract is awarded directly by the Bureau to the entity performing the work. No funds flow between CIEA and Compass Global Magazine. The two-contract structure is intentional: it lets each entity be paid for what it actually does, and it removes any related-party transaction between the nonprofit and the publisher. The findings are consolidated into a single Bureau-owned final report.

05What the Bureau Receives — Q1 2027

Feasibility report — written, EDB-owned, suitable for internal policy use
Demand dataset — manufacturer, worker, and founder registration data, structured
Manufacturer engagement readout — commitment depth from the December convening
Peer-hub comparative scan — what comparable institutions actually cost and produced
Strategy memo — three scenario paths for 2027, with cost and risk profile
Raw research artifacts — surveys, interview notes, instrumentation, transferable to any successor

06Why This Is Low-Risk for the Bureau

If the data is strong

The Bureau has cover, evidence, and a partner ready for 2027

EDB can proceed to formal AWS / NVIDIA / AMD MOUs, university anchor agreements, and a Phase 1 launch conversation with documented manufacturer demand to point to. The Bureau leads the next phase.

If the data is thin

The Bureau still has a clean, actionable industry dataset

No hub commitment was ever made. The research itself is useful policy input — manufacturer pain points, workforce gaps, infrastructure read — that the Bureau can apply to other industry programs. The 2027 hub conversation does not happen.

Bottom Line The Bureau is being asked to commission research, not to endorse a hub. The deliverable is data the Bureau owns and can use under either outcome — strong signal or weak — with no obligation to a 2027 follow-on. The two-contract structure keeps the nonprofit and the publisher operationally separate and removes any related-party complication. This is the most disciplined and the lowest-risk way to test the thesis before anyone spends real capital.