A · Team and IP
University ties, employment duties, inventors and applicants, code and design ownership. CityU Tech 300 eligibility must be confirmed in writing. Being a CityU P60 applicant is not, by itself, proof of Tech 300 eligibility.
HK–SZ FUNDING MAP · NOT AN AWARD LIST · UPDATED 2026-09-09
Maximise subsidies by splitting work, not adding ceilings
Pass eligibility gates first. Split R&D tasks and costs. Then file by concept, talent, corporate R&D, public trial, market and manufacturing. Awards depend on review and compliance. Published maxima are not PipeGuardian grants.
OPERATING IDEA
Treat a filing as a mirror of the R&D
Every application should point at a concrete G0–G6 work package. Tasks, people and a separate ledger come before the form. Adding poster maxima is not a financial plan.
Hong Kong is the natural home of auditable research evidence: concept validation, university commercialisation, corporate R&D matching, public-sector trials. Shenzhen / Qianhai amplify results that have already happened in Hong Kong: youth-project matching, real operations and R&D-spend support. Shared invoices, hours or deliverables across both sides create claw-back risk larger than the extra dollars.
PipeGuardian’s product shape — collision-tolerant inspection, well-head recovery, segment evidence — maps onto technology that can be trialled, accepted and written into a public-sector task. PSTS wants a real user and acceptance tests, not a lab selfie. This page therefore uses the same milestone language as the evidence gates and the roadmap.
FOUR GATES
Make the base real before you file
An unclear eligibility gate can mean refusal, delayed draw-down or later claw-back.
B · Hong Kong R&D substance
Company, business registration, people, equipment, contracts, accounts and project control that actually occur in Hong Kong. ESS, TSSSU and PSTS look for “mainly developed in Hong Kong”, not an email suffix.
C · Shenzhen / Qianhai substance
Office, staff, engineering, contracts, tax and youth shareholding / responsible-person tests where required. Matching awards presuppose a Hong Kong grant already won. Shenzhen cannot be used to tell the Hong Kong story in reverse.
D · No double funding
Separate project IDs, tasks, budgets, timesheets, invoices, payments and acceptance. The same cost is not reimbursed twice. A mechanical prototype, algorithm salary or rig hire must name which work package it belongs to.
HONG KONG LADDER
Hong Kong funding ladder
Figures below are public ceilings or rule summaries from official pages. They do not mean PipeGuardian has received, or will receive, any of them. Amounts, windows and eligibility follow the latest official notice.
| Scheme | Published support | PipeGuardian use | Key eligibility gate |
|---|---|---|---|
| HK Tech 300 Seed | HK$100,000 per team; 6–12 months. Round 21 opened 2026-09-23, closes 2026-10-07 | Customer discovery, Alpha, G0–G2 | Confirm CityU student / alumni / researcher or a CityU technology relationship |
| HK Tech 300 Angel | Equity investment up to HK$1 million | Beta, team and scenes | Hong Kong company and business registration, business plan, selection |
| HKSTP Ideation | One year, up to HK$100,000 | Concept validation and founder support | Project and team review |
| HKSTP Incubation | Up to HK$1.29 million | Ongoing R&D, commercialisation and park resources | Technology company, review and milestones |
| TSSSU-O / TSSSU+ | Up to HK$1.5 million per startup per scheme per year, up to 3 years | University technology commercialisation; TSSSU+ after private match | University relationship and internal selection; TSSSU+ needs private investment match |
| ESS | Government and company 1:1, up to HK$10 million per project | Core localisation, fusion and data R&D | Hong Kong company, business registration, matching cash and Hong Kong R&D substance |
| RTH-TC | Up to 4 people, up to 36 months; monthly stipend by degree band | Hire non-shareholder R&D engineers | Founders, shareholders and management cannot be the funded talent |
| PSTS-TC | Up to HK$1 million, usually ≤12 months | Controlled public-sector trial | Company owns the result, mainly developed in Hong Kong, real trial user and acceptance |
| BUD | Cumulative HK$7 million; up to 70 projects; Easy BUD up to HK$150,000 per item | Certification, brand, Mainland / overseas market | Market and business-upgrade spend; does not replace core R&D |
| NITTP | 1:1; up to HK$250,000 per company per financial year | External professional training for a mature team | Hong Kong registered company, eligible course and trainees |
SHENZHEN & QIANHAI
Hong Kong produces the evidence; Qianhai scales it
Shenzhen policy weights real operations, project landing and R&D spend. Do not manufacture related-party trades or duplicate costs for a match.
GBA-youth project 1:1 match
Eligible projects already funded by the Hong Kong government, a statutory body or a university may receive a one-off 1:1 award, up to RMB 1 million. Needs Qianhai operations, responsible-person shareholding and a no-double-funding check.
Shenzhen “Kunpeng Youth”
The 2026 filing window was 08-03 to 09-04. Check entity, team, real operations and R&D spend now. If this round does not fit, prepare the next. A closed window is not project failure; it only means this round’s eligibility did not line up.
Enterprise R&D subsidy
Support after real R&D spend, not seed capital. Start timesheets, materials, depreciation and a subsidiary ledger now.
Underground-pipeline data standard
PG-Atlas maps Shenzhen underground-pipeline coding, space and 3D information from the first edition, so a later government / operator interface has a place to land. Data-standard work can start before a flight score exists.
12-WEEK WAR ROOM
Turn filing into an operating system
A named output every two weeks. Every application draws from the same fact base.
W1–2 · Eligibility audit
Identity, school, company, equity, IP, Hong Kong / Shenzhen substance.
W3–4 · Work packages and budget
Split G0–G5 into non-overlapping tasks. Assign people, equipment and deliverables.
W5–6 · Technical evidence
Interviews, architecture, competitors, rigs, risk, plan and quotations.
W7–8 · Scene cooperation
University, inspection-company, park / public-sector letters of intent and acceptance metrics.
W9–10 · Finance and compliance
Cashflow, matching funds, procurement, related-party trades, IP and data.
W11–12 · Red team and submit
Challenge the pack on technology, market, team, HK–SZ substance, budget and risk.
Eligibility risk: current materials only record that a founder is a CityU P60 applicant. That sentence does not prove Tech 300 eligibility. Confirm in writing with CityU or the programme office before filing. Every amount on this page is a published-ceiling summary. It is not an award promise, and not a grant PipeGuardian has already received.
WHAT REVIEWERS ACTUALLY READ
Reviewers check story consistency
HK–SZ forms differ. They punish the same fault: website, business plan and application fighting each other.
If the site already wrote unverified endurance and customer counts, a PSTS trial plan that says “field verification is not complete” will be asked which sentence is true. If ESS writes localisation as Hong Kong core R&D, and a Shenzhen invoice books the same engineers’ wages in full, audit will ask where the work happened. If Tech 300 materials stress a CityU relationship and the team page has no verifiable eligibility path, the window closes before the deadline.
The war-room product is therefore not “an Excel that sums every published maximum”. It is an eligibility memo, non-overlapping work packages, an evidence-gate table, a scene letter-of-intent template, and a related-party list. Amounts only mean something after those files exist. Public ceilings change. Cite the official page and record the retrieval date.
WORK PACKAGE EXAMPLES
The same money cannot buy two stories
The split below is a rehearsal, not a submitted budget. The point is to practise non-overlap before anyone writes an application.
A concept-validation pack covers interviews, scene cards, a P0 envelope and a G0 report — Seed / Ideation scale. A prototype pack covers cage tooling, a collision rig and a G1 failure list — incubation or TSSSU process files. A core-algorithm pack covers localisation degradation, time sync and a data dictionary — ESS-class schemes that demand Hong Kong R&D substance. A trial pack covers task cards, a contrast procedure and acceptance metrics — PSTS. A market pack covers certification advice and brand — BUD, and it explicitly does not buy core R&D.
If the prototype pack already booked the collision rig, the algorithm pack cannot list the same hire as primary equipment. If the trial pack promises “a real public-sector task”, the website cannot also say “we have not entered any field”. Red-team week prints those sentences on one sheet. Amounts are filled after that, below the published official ceiling, and they keep an “unawarded” label.
Windows expire. Eligibility does not auto-extend. Tech 300 Round 21 dates follow the official page. If the 2026 Kunpeng Youth window has closed, move the pack to the next round; do not rewrite history. The filing-room calendar is internal. The site does not manufacture an “award imminent” expectation.