Who Controls India’s Drones?

India flies drones it cannot control

India is building drones faster than it can control them. The airframes, motors and even the sensors are increasingly assembled at home, but the component that decides where a drone goes and what it does — the flight controller, and the autonomy software layered on top of it — remains substantially foreign, and predominantly Chinese. If the companion report asked who builds India's drones, this one asks the sharper question: who controls them? The brain of the drone is imported, and that is a national-security exposure and a domestic market opportunity at the same time.

The 39% that is really 70-80%

On 3 April 2025 the Government of India told the Lok Sabha that 39% of flight controllers in DGCA-certified small drones are sourced from China. That figure is real but narrow: it counts only certified small drones and excludes the nano and micro categories that dominate by volume, the uncertified and grey-market fleet, aftermarket boards, and fully imported platforms such as DJI. Add those back and effective dependence is 70–80%. The most acute exposure is deeper still: flight-control processors and MEMS inertial sensors are approximately 100% imported, so even “indigenous” controllers are assembled on foreign silicon.

Where the control stack is exposed

The control stack is a layered system, and value and vulnerability both concentrate at the ends — in the silicon at the base and the certified, electronic-warfare-resilient autonomy at the top. India captures value in assembly, firmware adaptation and, increasingly, mission software, but leaks it in the two layers that matter most strategically. Operation Sindoor in May 2025 made the point operationally: imported flight controllers lack the anti-jam and anti-spoof resilience that contested environments demand, turning electronic-warfare resilience and GPS-denied navigation into the highest-priority capabilities — and the ones where indigenous firms can differentiate.

Four frameworks

The report turns a diffuse sense of “import dependence” into a ranked, decision-grade picture using four proprietary frameworks: the Drone Autonomy Value Chain (where India captures and leaks value, layer by layer), the Strategic Vulnerability Index (every layer scored on dependence and strategic importance), the Import Vulnerability Matrix (which exposures justify state investment versus ordinary localisation), and the India Drone Autonomy Maturity Model (five levels, and why value migrates toward exactly the capabilities India is weakest in). Together they explain why a localisation programme optimised for the easy layers can raise the headline percentage while leaving every Critical exposure intact.

Where to build now

Policy has converted localisation from a preference into a condition of market access: the new INR 2,000 crore PLI scheme mandates 40% domestic value addition in flight controllers, navigation and communication modules by FY2027-28. Against a market growing from USD 150–200 million in 2026 toward USD 800 million–1.2 billion by 2035, that defines a domestic substitution opportunity above USD 500 million in addressable annual value. Opportunity sizing — not market sizing — locates where capital earns a return.

Addressable values are Techadyant modelled estimates for the early 2030s.

Winners and losers

Four levers — the PLI scheme, the Army's 515 Base Workshop validation model, BVLOS commercialisation and indigenous-content procurement preference — are about to redistribute value. The clear winners are vertically integrated indigenous players (ideaForge, listed since 2023), patented flight-control startups (Zuppa), electronic-warfare and GPS-denied specialists (Tsalla), Army-validated entrants (Yaanendriya) and defence primes (BEL, Raphe mPhibr). The exposed are OEMs that rely on imported Pixhawk-class boards without localising, Pixhawk-clone importers, uncertified “AI” software claimants and Chinese suppliers. Beneath the actor-by-actor map sits one structural shift: capital is concentrating at the platform layer while the flight-control layer — where strategic value is highest — is starved.

A path to 2035

The realistic doctrine is selective sovereignty. Build now what is ready — indigenous flight-control hardware, EW-resilient and GPS-denied autonomy, AI mission software, and certification and integration services. Seed the hard layers in parallel: a RISC-V flight-control system-on-chip and MEMS packaging are five-to-eight-year undertakings, and processors should be exempted from the 40% localisation count until domestic capacity exists, or the target is unreachable. Success is one measurable number: import dependence moving from roughly 70% toward 45% by 2030.

What the full report adds

This online edition gives you the thesis, the import-dependence finding, the vulnerability stack and the Build-now map. The complete edition adds the full Strategic Vulnerability Index across the control stack, the Drone Autonomy Opportunity Map and maturity model, the Winners & Losers grid, five costed recommendations with owner, cost and timeline, an evidence-hierarchy methodology, and twelve deep-dive chapters on flight-controller technology, AI and autonomy systems, market sizing and segmentation, the import-dependence and PLI analysis, the regulatory landscape, the competitive landscape, investment and funding, and the technology roadmap to 2035 — seventeen chapters and thirty-five figures in all.