The Guangdong-Hong Kong-Macao Greater Bay Area has no shortage of innovation assets. It has researchers, engineers, hospitals, industrial clusters, capital, public-sector users, and an increasingly sophisticated base of technology firms. Yet cross-boundary innovation still too often begins with an organization looking for something to sell, rather than a defined problem looking for the right people to solve it.

That distinction matters. The usual route asks a team to form a company, raise capital, build a product, and then search for a market. It can produce outstanding enterprises, but it is not the only route to innovation. It’s costly and slow for problems that are shared across a region. It also excludes many capable people — engineers, clinicians, teachers, operations specialists and researchers — who can solve a real problem but do not wish to become founders, fundraisers, and full-time company managers.

Hong Kong’s first five-year plan provides a timely opening. It calls for Hong Kong and Shenzhen to pursue “collaborative research and development, coordinated transformation and shared development”, with the Hetao Shenzhen-Hong Kong Science and Technology Innovation Cooperation Zone as a focal point. The next step should be to give this ambition an operating system: a cross-boundary mechanism built around four forms of sharing — shared problems, shared funding, trusted data, and shared accountability.

Start with a shared problem, not a generic call for solutions

A public problem bank, jointly governed by relevant institutions on both sides of the boundary, would be the first building block. Government departments, hospitals, universities, firms and community organizations could submit defined challenges: reducing handover delays in cross-boundary logistics; improving access to safe digital services for older adults; or testing low-carbon equipment in different operating environments.

Each challenge should identify its problem owner, intended users, delivery boundaries, success measures, available resources and route to adoption. That prevents innovation from becoming a technology demonstration in search of a customer. It also makes participation more open. A university laboratory, a corporate innovation team, a professional association, or a temporary cross-disciplinary group assembled in Hong Kong and Shenzhen should be able to compete to solve a challenge. A successful solution may later become a startup or a commercial product, but incorporation should not be the ticket of entry.

Fund the work and reward the outcome

The second building block is shared finance. A joint mechanism should have two pools. An execution pool would cover necessary costs of discovery, prototyping, testing, compliance, and field deployment. An outcome pool would reward verified performance through prizes, follow-on orders, or adoption opportunities.

Either pool on its own is inadequate. Reward only success, and the promising teams may run out of resources before a pilot program is complete. Fund only execution, and the program risks becoming just another grant program that pays for reports rather than results. Funding should follow a jointly agreed problem and transparent milestones, not institutional prestige or the polish of a presentation. This is how public money can move from supporting organizations in the abstract to purchasing measurable solutions.

Treat data as a governed resource, not a prize to be collected

The third building block is trusted data. Many cross-boundary projects fail not because the underlying technology is unavailable but because data cannot be accessed, formats are incompatible, or permissions are unclear. The answer is not unrestricted data flows. It is project-specific, legally compliant data governance.

For each challenge, participants should know the minimum data they may use, for what purpose, for how long, through which interface, and under what security controls. Requirements for anonymization, storage, access logs, audit trails and deletion should be agreed upon before a pilot program begins. Data should be accessed in controlled environments, not handed over without restraint. In this model, compliance does not arrive at the end as a legal obstacle; it is part of the solution design from day one.

This is an area in which Hong Kong can add distinctive value. Under the “one country, two systems” framework, its common-law tradition, international professional services and experience in finance, data governance and dispute resolution can help translate requirements into trusted interfaces. Shenzhen contributes a deep industrial base, engineering speed, and a rich range of application settings. The value lies not in either city copying the other but in making their strengths work together.

Share accountability, including for a well-run failure

The fourth building block is shared accountability. Too many partnerships explain how success will be celebrated, but not what happens if an experiment fails, data proves insufficient, or an intended user cannot adopt the solution. Before a project starts, the parties should agree on risk sharing, intellectual property, procurement conditions, exit terms, and a process for reviewing failure.

The problem owner should be accountable for defining the need and providing a viable test setting. The solution team should be accountable for technical quality and the honest disclosure of risks. Funders and evaluators should be accountable for milestones and decisions to continue, adapt or stop. A well-run failure should not be treated as waste. If the process was compliant and the evidence is retained, its lessons can return to the problem bank so that the next team does not begin from zero.

Every challenge also needs an adoption owner. A system that publishes problems and funds pilot programs but cannot identify who will validate, procure, license or scale a successful solution will become another competition. The adoption route must be designed before development begins. This need not begin as a huge new bureaucracy. The Greater Bay Area could start with a limited number of challenges in which demand is clear, data boundaries are manageable, and results can be measured: logistics, low-carbon operations, or selected public services. Each pilot program should have one joint governance group, one set of metrics, and one end-to-end owner. Its scorecard should track time to solution, adoption, avoided duplication and measurable benefits — not simply memorandums signed or events held.

The Greater Bay Area’s next advantage will not come from having one more forum or one more incubation program. It will come from turning separate pools of talent, capital, data and responsibility into a system that can solve shared problems repeatedly. Shared problems bring the right people in. Shared finance keeps the work moving. Trusted data makes results testable. Shared accountability makes collaboration credible. That is how cross-boundary cooperation can move from good intentions to a productive, repeatable capability.

The author is founding chairman of the Hong Kong Innovative Technology Development Association.

The views do not necessarily reflect those of China Daily.