Singapore is betting on AI to keep its manufacturing base competitive. For most SMEs, the question isn’t whether to take part, but how to turn national ambition into one better decision on the shop floor.
Singapore has never treated manufacturing as a sunset industry. While many advanced economies let their factories drift offshore, Singapore set out a Manufacturing 2030 vision to grow manufacturing value-add by 50% over the decade, and has kept manufacturing at around a fifth of the economy. Now the country is making a second, related bet: that artificial intelligence will decide which manufacturing economies stay competitive, and that Singapore intends to be one of them.
For the multinationals in Tuas, Jurong and Woodlands, this is familiar territory. They have global AI teams and lighthouse plants. The real question is what national ambition means for the 50-person precision engineering firm, the food processor in Senoko, or the moulder supplying the semiconductor cluster. Those firms are the 99%, and the national strategy only succeeds if they come along.
The framework has been building for several years:
The direction is clear. The details of each mission and programme will keep evolving, so treat the specifics as moving parts and the direction as fixed.
Here is the uncomfortable arithmetic. A national AI strategy measured only by what multinationals do will look impressive and change little for most of the economy. The multinationals would have adopted AI anyway. The national value lies in lifting the supply base around them: the Tier-2 and Tier-3 firms whose quality, reliability and resilience those multinationals depend on.
When those firms can connect their data, prove their outcomes, answer a carbon request and decide well under pressure, everyone benefits. Multinationals get fewer disruptions and a supply base that can meet the AI, cyber, and sustainability standards they now demand of their chain. The SME moves from price-taker toward preferred supplier. And Singapore keeps a manufacturing ecosystem that is genuinely hard to replicate elsewhere.
That is why I believe the success of the national AI strategy in manufacturing will be decided not in the lighthouse plants, but in the 99%.
National programmes carry a quiet risk for SMEs. When incentives reward visible “AI adoption”, firms are pushed to buy the top of the technology stack before the foundation exists. An AI quality system built on disconnected, undefined, and unowned data will produce outputs the floor doesn’t trust. Trust collapses, the spend is written off, and the next, sounder attempt starts with a sceptical team.
I have watched this pattern repeat through several waves of Industry 4.0 support. The firms that benefited most were not those that bought the most advanced systems. They were those that used support to build the unglamorous layers first: connected data, shared definitions, captured knowledge and a clear owner for each important decision.
If you lead an SME, here is how I would translate national ambition into practical steps.
1. Start with one decision, not one grant. Before you apply for anything, name the recurring decision that most affects a customer outcome: when to change a tool, whether a batch is fit to release, which order to run next. Support schemes are far easier to use well when you know exactly what you are trying to improve.
2. Use assessment before investment. SIRI and similar assessments help you locate the missing layer. Read them as a diagnosis, not a shopping list.
3. Fund the foundation first. Connectivity, open data standards and knowledge capture are rarely glamorous, but they are what make every later AI investment pay. If a proposal asks you to fund the roof before the foundation, push back.
4. Build people capability alongside technology. The binding constraint in most SMEs is not equipment; it is management capability and the capacity to absorb change. Training for your leadership team is often the highest-leverage use of support.
5. Think supply chain, not just factory. Your multinational customers are already asking about AI governance, cybersecurity and carbon. The national push and their supplier requirements point in the same direction. Answer both with the same work.
For policymakers and the bodies that serve SMEs, I would offer three suggestions drawn from the book’s Policy Lens:
I am optimistic. Singapore’s strengths, trusted institutions, a dense manufacturing ecosystem and a government willing to act are exactly what the 99% need. Over the coming months, I am working with partners across Singapore’s SME support ecosystem to bring a practical, funded programme for SME leadership teams that want to do this properly: one decision, ninety days, proven against a baseline.
If you would like to be among the first firms to take part, join the waitlist below. And whether or not you do, ask yourself this week: if the national AI strategy succeeds, what will your factory be doing differently in three years? If you cannot answer yet, start with one decision.
Your next step: Funded programme waitlist