Agriculture & Forestry

Neither use case advanced as a coordinated MOSTI programme with a named owner, budget line and public milestones. What happened instead is more interesting: the underlying problems were real and urgent enough that industry moved on its own, largely outside the roadmap's framing.

The driver was a labour crisis, not a policy push. The pandemic-era freeze on migrant labour left plantations critically short-handed; one widely-cited industry estimate put unharvested fruit losses at around RM10.46 billion in just the first five months of 2022.^[1] That economic pressure — far more than the roadmap — is what pulled capital into agricultural automation.

The harvesting use case became a corporate R&D race

The big plantation groups built their own robotics capability. Sime Darby Plantation (since rebranded SD Guthrie) established a dedicated robotics unit and aimed to eliminate manual labour from all non-harvesting tasks by end-2023, targeting a 55% cut in plantation headcount and a land-to-man ratio of 1:17.5 hectares.^[2] Local robotics firm Meraque supplied drone and autonomous-vehicle systems to FGV, Boustead and Sime Darby; IOI Corp and others scaled drone spraying and mapping.^[3] By 2024, US robotics firm EarthSense was running autonomous fertiliser-spreader trials on SD Guthrie estates.^[4]

Crucially, the hardest part of the roadmap's vision — fully autonomous FFB cutting on tall palms — remained unsolved. A 2024 MPOB review still framed it as an open problem across six unsolved technologies, from ripeness detection to tree-climbing.^[5] The government and industry committed roughly RM60 million to automated-harvesting R&D, but commercial-scale autonomous harvesting stayed years away.^[3]

The supply-chain use case reappeared as traceability, not procurement

The roadmap framed Project 1 around procurement and inventory optimisation. What Malaysia actually built was a supply-chain traceability system — driven less by efficiency and more by the EU Deforestation Regulation (EUDR). The eMSPO platform now manages the full Malaysian Sustainable Palm Oil certification lifecycle and connects plantation, mill and export into a single traceable chain.^[6] The same destination the roadmap pointed at — data flowing across the value chain — but reached through a regulatory door, not an AI-procurement one.

The institutional context has shifted underneath this sector. The roadmap was a MOSTI publication, but national AI coordination has since moved toward the Ministry of Digital and the National AI Office. Agricultural automation, meanwhile, sits largely with the plantation industry and the commodities ministry — meaning the original "agriculture as a National AI Use Case" framing has effectively dissolved into sector-specific programmes that no longer reference the roadmap.

On traceability, Malaysia is close to its destination: MSPO certification covers roughly 90% of planted area as of 2025–2026, and an integrated National Traceability System (Sistem Ketertelusuran Nasional) consolidating existing databases is targeted for full implementation in March 2026.^[7] MSPO 2.0 (MS 2530:2022) became mandatory in January 2025 with stricter traceability requirements.^[8]

On harvesting, the honest status is "real progress, no finish line." Mechanisation and drones are now routine on large estates; full autonomous harvesting is still a bet the industry is funding rather than a capability it has shipped. Industry leaders themselves describe it in the language of placing bets, not declaring victory.^[4]

The agriculture story is the roadmap's pattern in miniature. The diagnosis was right — labour scarcity, supply-chain opacity, a globally significant industry exposed to disruption. But the delivery happened despite the roadmap rather than through it: corporate capital and external regulation, not a coordinated national AI programme, did the work. This is exactly the gap a coordinating body was supposed to close — see the governance body that never formed.

For the 2026–2030 cycle, three things are worth watching: whether the next national AI plan re-adopts agriculture as a named priority or leaves it to industry; whether the National Traceability System actually ships on its March 2026 target and becomes a genuine data backbone; and whether anyone solves autonomous FFB cutting at commercial scale — the single capability that would convert today's partial mechanisation into the integrated system the roadmap originally described.