Singapore’s automation trade body has introduced a two-year, use-case-specific recognition scheme for factory-floor AI. Benchmarked against ISO/IEC 42001, the framework aims to close the region’s governance gap ahead of its public registry launch in late 2026.
AutomationSG, Singapore’s trade association for the automation and robotics sector, launched a recognition scheme for industrial artificial intelligence on Wednesday (29 July).
The Trusted Industrial AI Ready (TIA-Ready) Framework assesses factory AI deployments on a use-case basis, awarding recognition valid for two years. The voluntary scheme arrives as AI reaches Southeast Asian factory floors faster than the governance required to manage it.
Advertisments
TIA-Ready at a Glance
- What it is: A voluntary, use-case-specific recognition scheme for industrial AI, run by AutomationSG
- Pathway: Three stages – Managed, Governed and Ready – over a seven-step process
- Benchmark: Aligned with ISO/IEC 42001, but not an ISO certification or statutory approval
- Validity: Two years, with an annual declaration that scope and risk have not materially changed
- What it isn’t: No certification of safety, performance or legality – liability stays with the recognised company
- Status: No recognitions yet; public registry targeted for Q4 2026
In a preliminary survey of 20 companies by AutomationSG, 55 percent had no formal control over AI tools, and 30 percent lacked accountability for AI. The gap extends well beyond this sample: a February 2026 McKinsey survey of 330 AI-using companies across six Southeast Asian economies, published with the Singapore Economic Development Board, found 41 percent had suffered negative consequences from AI inaccuracy, and 21 percent had experienced cybersecurity incidents.
AutomationSG’s chief executive, Saw Biing Huei, argued that industrial AI governance must precede the technology itself. Trusted AI, he noted, “does not start with the dataset or the test result”, but with management deciding the AI’s purpose, its strict limitations, whose decisions it may influence, and what risk the company is willing to accept.
Benchmarks and Current Status
Targeted at companies deploying AI in industrial automation, robotics, manufacturing, and operational technology, TIA-Ready requires a company to define a single use case, its operating boundary, and its risk profile before moving through three stages: Managed, Governed, and Ready.
Formerly the Singapore Industrial Automation Association (SIAA), AutomationSG benchmarks the scheme against ISO/IEC 42001, the international standard for AI management systems. However, the body explicitly states that TIA-Ready is not an ISO certification, statutory approval, or a legal certificate of compliance. Full ISO certification remains rare in the manufacturing sector; industry tallies put the global total at roughly 350 organisations by spring 2026, most of which are technology firms.
Currently, no company holds the TIA-Ready recognition. The pathway’s final step is a listing on a public registry targeted for the fourth quarter of 2026. Thus far, validation rests on two Southeast Asian SME pilots – Singapore machine vision integrator JM VisTec System and Malaysian engineering software provider CADVision Systems – which ran desktop validation rather than live deployments.
How TIA-Ready Works
The recognition process follows seven steps: from initial evaluation and scope definition through governance implementation, evidence preparation, independent assessment, an AutomationSG decision, and ultimately, the registry listing.
Assessors verify that evidence exists, is complete, is traceable to framework requirements, and that governance is actively implemented. Recognition is valid for two years and requires an annual declaration that the scope and risk profile have not materially changed, followed by a reassessment at the end of the term.
The boundaries of the scheme are explicit: recognition confirms only the existence and traceability of required evidence. It does not certify the safety, performance, legality, or fitness of the AI system, and liability remains strictly with the recognised company. Additionally, AutomationSG’s published materials do not currently state participation costs or clarify whether companies outside Singapore are eligible to seek recognition.
A Crowded Policy Landscape
The framework was developed over four months by a steering committee that included the National Robotics Programme, Siemens, Infineon Technologies, STMicroelectronics, Beckhoff Automation, TÜV SÜD, Cantier Systems, and HOPE Technik. It was launched at the Automation SolutionGO! x Regional Industry Networking Conference 2026 at Singapore Polytechnic, witnessed by Alvin Tan, Minister of State at the Ministry of Foreign Affairs and the Ministry of National Development.
This launch caps a busy fortnight for Singaporean AI policy. On 20 July, the country’s Infocomm Media Development Authority (IMDA) released transparency guidelines for generative AI chatbots – described as among the first globally – alongside guidance on digital twins and federated learning. Expanding on this push for education, Singapore Polytechnic will run the region’s first Industrial AI and Governance Literacy programme beginning in October.
Image credit: AutomationSG
Discussion about this post