This article first appeared in The Edge Malaysia Weekly on June 1, 2026 - June 7, 2026

For generations, the art of agriculture has been governed by intuition. Farmers relied on historical weather patterns and “the skilled human eye” to cultivate life from the soil. Today, however, that intuition is colliding with a harsh mathematical reality.

According to the United Nations’ Food and Agriculture Organization, the global population will reach 9.1 billion by 2050, requiring a massive 70% increase in global food production. We are already falling behind: the 2026 Global Report on Food Crises (GRFC) reveals that about 266 million people across 47 countries face high levels of acute food insecurity due to compounding systemic shocks like climate extremes and supply chain disruptions.

Here in Malaysia, the crisis is a sovereign one. Our national food import bill continues to hover near an alarming RM80 billion annually, exposing the domestic economy to severe currency fluctuations. This deficit is driven by a stark land utilisation imbalance: despite abundant arable land, less than 10% is dedicated to food crops, while over 90% is occupied by industrial cash crops like oil palm and rubber.

To survive the demographic and climatic cliff ahead, the agricultural sector must urgently transition from a reactive, intuition-based practice to a proactive, data-driven ecosystem. We must move towards Agriculture 5.0, leveraging Fourth Industrial Revolution (IR4.0) deep tech to unlock both unprecedented business success and genuine environmental, social and governance (ESG) leadership.

1. Farming in the dark

Farming has always been a bet against the weather. For generations, that bet was calculated. Farmers read the seasons, trusted their soil and planned around weather patterns that held relatively true year after year. Climate change has broken that contract entirely. Droughts arrive unannounced, pest migrations accelerate and by the time anyone notices something is wrong, the damage is already done.

There is another problem that rarely gets talked about, the sheer size of a modern farm makes it impossible to watch closely. Human scouts walk the perimeters. But what is happening in the middle of a 200ha plantation? Nobody really knows, not until it’s too late.

The deep tech AI solution

This is where the convergence of satellite imaging, multispectral drones and machine learning starts to feel less like technology and more like a superpower. Rather than waiting for a yellowing leaf to signal distress, artificial intelligence (AI) systems now detect cellular-level stress in plants, water deficiency and early signature of a fungal infection. All this is a full 10 to 14 days before any human eye can spot it. The system does this by continuously synthesising data from soil chemistry, hyper-local weather feeds and drone-captured imagery processed through normalised difference vegetation index (NDVI) algorithms. It is essentially giving farmers a window into the future.

The outcomes are measurable. Farms using these predictive models are consistently seeing 5% to 10% increases in marketable yields. Not by farming more land, but by farming smarter within the same land.

In Malaysia, this is already moving beyond pilot projects. Agritix deploys satellite and multispectral imagery to flag plantation anomalies in real time. Bit Group’s AI Engine connects weather and soil data to live yield predictions and disease alerts. And the government-backed MyAgriTECH initiative is systematically rolling out AI-driven environmental analysis across the sector, reporting over 90% efficiency gains in resource management.

2. The true cost of ‘spray and pray’

Agriculture is quietly the world’s biggest water consumer. And for decades, the standard approach to protecting crops has been blanket applications of fertilisers and pesticides across entire estates, regardless of whether every corner actually needs it or not. It’s expensive, ecologically destructive and increasingly incompatible with the ESG expectations now being placed on the sector.

The deeper damage is less visible but just as serious. This kind of indiscriminate application destroys the soil microbiomes that keep farmland fertile over the long run. Chemical run-off poisons surrounding waterways. And as input costs keep rising, farms doubling down on this approach are essentially paying more to cause more damage, with diminishing returns at every step.

The deep tech AI solution

The smarter alternative is not to use less by guessing, it is to use exactly what is needed, exactly where it is needed.

Internet of Things (IoT) sensor networks now make that precision possible at scale. Buried across a farm, these low-cost sensors feed continuous readings on soil moisture, pH and nutrient levels into an AI that responds in real time. If a specific zone runs dry, the system calculates the exact water volume required and opens the right valve. No human decision needed, no water wasted elsewhere.

Pest management has become even more sophisticated. Deploying scouting drones equipped with multispectral sensors fly the fields, identifying cellular stress signatures from pest activity long before any visible damage appears. Once a hotspot is flagged, AI processes the imagery and diagnoses the exact pest species using convolutional neural networks, then generates a precise “prescription map”. Heavy-payload spraying drones then fly those coordinates autonomously, opening nozzles only over confirmed problem zones.

The results speak for themselves. Pesticide use drops by 35% to 60% and water consumption falls by up to 90% compared with conventional methods. This is not just good for the environment, it directly cuts input costs and strengthens a farm’s sustainability credentials.

Locally, Terra Drone Agri is scaling smart plantation operations through advanced aerial mapping. iTani’s AI-powered laser recognition system has helped orchard farmers cut pesticide use by as much as 90%. The Blanket Spraying Hybrid Drone has effectively replaced 20 manual labourers with a two-person team while achieving 100% spray accuracy.

3: Who will work the fields tomorrow?

There is an uncomfortable statistic sitting at the heart of global agriculture: in Japan, the average farmworker is 67.6 years old. The US quietly imports nearly half a million temporary labourers every year just to keep its harvests from failing. The workforce that weeds, maintains and harvests is ageing out faster than it can be replaced and the food systems built around that labour are now genuinely fragile.

The vulnerability runs deeper than just numbers. When the harvest depends almost entirely on human hands, every disruption — a pandemic, a shift in border policy, an unusually hot season — can cascade into a food supply failure. The margin for error is shrinking every year.

The deep tech AI solution

Robots do not get tired, they do not age. The drone swarm — three to five fully autonomous units mapping and spraying in coordinated formation — is now a reality. Smart ground rovers navigate crop rows with machine vision that is precise enough to identify and target individual weeds in milliseconds, reducing chemical use by up to 90%. And robotic arms fitted with soft grippers and deep learning models can now assess fruit ripeness visually and harvest them. Something that, not long ago, seemed like science fiction.

Malaysia is among the countries moving most deliberately to close this gap. The Harvest-Mate Robot applies machine learning and mechanical arms to oil palm harvesting, boosting productivity by over 120% compared with manual methods. Poladrone deploys autonomous drones for targeted weed spraying, eliminating not just the labour but also the genuine hazard of workers carrying heavy chemical backpacks through dense, uneven terrain.

The shift is not about replacing farmers. It is about ensuring the fields do not go unworked when the people who once tended them are no longer there.

Transitioning to Agriculture 5.0 requires more than just available technology, it requires a coordinated national ecosystem that bridges the gap between deep tech developers and traditional farmers. The most critical breakthrough for everyday farmers is the Drone-as-a-Service (DaaS) model. Smallholders do not need to buy an expensive spraying drone. Instead, they hire localised drone operators for a flat fee, completely removing the capital expenditure (capex) barrier and democratising access to precision agriculture. Platforms like the Agrimor SuperApp are already connecting farmers to on-demand drone services, driving up to 67% increases in crop yields and 50% reduction in food production costs.

In Malaysia, this mandate is aggressively driven by the Malaysian Research Accelerator for Technology and Innovation (MRANTI). Recognising that national food security cannot rely solely on massive corporate estates, MRANTI focuses on digital inclusion for smallholder farmers.

For a commercial farmer or tech entrepreneur, deep tech is only viable if the math works. The hesitation to adopt Agriculture 5.0 often stems from the perceived cost of entry, but isolating capex against operational expenditure reveals a highly lucrative reality.

The data reveals a different reality. For a commercial farm in Malaysia, an upfront investment of RM50,000 to RM100,000 in drone ecosystems and AI sensors is not a sunk cost, it is a rapid-return asset. By reducing volatile input costs, slashing fertiliser waste by 20% and pesticide use by up to 50%, the technology generates a breakeven point within just one to two harvest cycles.

The transition to Agriculture 5.0 is no longer an imported, futuristic concept. Through the Malaysia Drone Technology Action Plan 2022-2030, MRANTI coordinates scalable solutions to address labour and yield challenges. They provide vital physical testbeds like Area 57, a Drone Tech Centre of Excellence where AI obstacle avoidance and drone swarming are stress-tested before deployment.

For the modern agricultural entrepreneur, an upfront investment in drones, AI and IoT is not a sunk cost, it is a rapid-return asset. Moving from intuition-based farming to a data-driven ecosystem is no longer just an environmental imperative, it is the most logical business decision we can make to secure our food supply and lead the future of global agriculture.

Akhil Gupta is deputy chair of the AI Chapter and AI Academy chair at PIKOM (National Tech Association of Malaysia). He is group CEO of Talbotiq Technologies and Total IT Global.

Save by subscribing to us for your print and/or digital copy.

P/S: The Edge is also available on Apple's App Store and Android's Google Play.