Defence D-HUMR
Defence D-HUMR is a defence-grade unmanned ground vehicle platform engineered with hardened compute, encrypted communications, and a ruggedised sensor suite for outdoor reconnaissance and logistics. Priced at ₹20,00,000 per unit, it represents the highest price point across DeepGrid's portfolio and serves as an early sovereign revenue stream independent of automotive mandates.
- Listed price
- ₹20,00,000
- FY2032 revenue
- ₹30 Cr
- Share of plan
- 2.7%
- Gross margin
- 68%
- First revenue
- FY2027
- Volume plan
- 150 @ FY32
From input to outcome
Proposed product architecture. The scene illustrates its application, not a shipped kit.
Surround sensing reads terrain, obstacles and people
SoC2 builds a local map and plans a traversable route
Autonomy drives, with a teleoperator able to take control
Mission payload does its job at the destination
Drag to rotate. Conceptual layout, not a mask.
On the die
The case in brief
- Defence D-HUMR provides sovereign validation and early revenue independent of commercial automotive mandates, scaling to ₹30.0 Cr at a 68% gross margin by FY2032.
- The ruggedised unmanned ground vehicle commands a ₹20.00 L unit price, leveraging SoC2 outward perimeter fusion to execute logistics and reconnaissance in hazardous terrain.
- The platform has already achieved initial external revenue validation on FPGA hardware, with commercial ramp pacing tied to multi-year defence procurement cycles.
What it is used for
Defence Logistics and Unmanned Reconnaissance
- Buyer
- Indian Army and MoD procurement programmes
- The job
- The military requires autonomous logistics and reconnaissance platforms for hazardous missions where sending an unmanned ground vehicle is preferable to risking human personnel. Operating environments often feature GPS-denied conditions and complex terrain requiring secure, encrypted operational integrity.
- What DeepGrid supplies
- DeepGrid delivers an unmanned ground vehicle featuring surround sensing, an autonomy stack running on SoC2 with military-grade encryption, autonomous navigation in GPS-denied environments, a teleoperation link, and a secure command post data relay.
Intelligent Perimeter Threat Fusion
- Buyer
- Defence/Paramilitary organisations (DRDO/MCEME pipeline)
- The job
- Perimeter security requires identifying coordinated breach threats across access boundaries without flooding security operators with nuisance false alarms that cause surveillance systems to be disabled.
- What DeepGrid supplies
- DeepGrid supplies an outward-facing multi-sensor fusion stack (proven via the DGrid Sentinel console) that correlates external vehicle tracks and fence-line human movements along a timed threat ladder, holding benign tracks while explaining escalation rationale in words.
The story
An unmanned ground vehicle for defence logistics and reconnaissance.
Defence D-HUMR is DeepGrid's highest unit-price line and the only portfolio product selling directly to government defence buyers. Built for outdoor logistics and reconnaissance tasks where deploying robotics is preferable to risking personnel, the system operates on multi-year procurement cycles rather than commercial automotive timelines. It serves as the portfolio's earliest external validation, having already secured initial revenue on FPGA hardware ahead of production silicon rollout.
- ₹20.00 LUnit price
- RoboticsSegment
- OutdoorOperating domain
- FY2027First revenue
- UGV platformSurround sensing component
- SoC2 computeAutonomy stack component
Outward-facing perimeter fusion across radar, thermal, and camera tracks.
The platform runs DeepGrid's core fusion stack outward along perimeters rather than forward along highways. Surround sensing captures terrain, obstacles, and personnel, while SoC2 classifies individual sensor tracks, builds a local traversable map, and correlates distinct events like access-road loitering and fence-line movement into unified threat assessments. Autonomy drives the vehicle with integrated teleoperation override, executing payload delivery while dependent primarily on sovereign procurement timelines rather than technical milestones.
- Sense, compute, decide, actSignal chain architecture
- Terrain, obstacles, peopleSurround sensing inputs
- Local map and routingSoC2 compute task
- Autonomous with teleoperation linkDrive execution
- Mission payload interfaceDestination action
- Slow, lumpy procurement cyclesKey dependency
Indian military procurement replacing personnel with autonomous systems in hazardous zones.
Purchasing is driven by the Indian Army and Ministry of Defence procurement programmes seeking unmanned systems for hazardous perimeter, reconnaissance, and transport missions. Channels include direct institutional acquisition and Government e-Marketplace (GeM) contracting frameworks. Although defence qualification timelines are slow and cycles are multi-year, secured platform incumbency creates exceptionally durable market defensibility against displacement.
- Indian Army and MoDTarget buyer
- Logistics and reconnaissance missionsBuying trigger
- GeM and government tendersChannel
- Indian Army robotics procurementDemand pool
- Existing Rs1 Cr revenueSource validation
A high-margin sovereign revenue stream independent of automotive mandates.
Defence D-HUMR scales from 6 units in FY2027 to 150 units in FY2032, generating ₹30.0 Cr in revenue at a 68% gross margin as management projections. While representing 2.7% of group revenue, its strategic importance lies in customer diversification through sovereign defence budgets that operate completely free of passenger vehicle regulations. Prior revenue earned on FPGA implementations provides early external commercial validation for the underlying architecture.
- 150 unitsFY2032 volume projection
- ₹30.0 CrFY2032 revenue projection
- 2.7%Group revenue share
- 68%Gross margin
- FPGA early revenuePortfolio role
- ₹1.2 CrFY276 units
- ₹3 CrFY2815 units
- ₹4.6 CrFY2923 units
- ₹9 CrFY3045 units
- ₹17.6 CrFY3188 units
- ₹30 CrFY32150 units
Intelligent threat correlation demonstrated through the DGrid Sentinel console.
Perimeter decision logic has been verified using the DGrid Sentinel demonstrator, a live console running the production perception stack against a simulated environment. The system fuses disparate camera, radar, and thermal inputs along an escalation ladder from loitering to approach and breach risk, handing tracks across ground units, drones, and maritime buoys. The demonstrator proves the decision layer in simulation, not a fielded deployment, emphasising false-alarm restraint by holding benign tracks without generating disruptive false alarms.
- DGrid SentinelPerimeter threat fusion
- Simulation, not a fielded deploymentValidation status
- Loiter, approach, breach riskTimed escalation ladder
- Vehicle and human tracksCorrelation mechanism
- Ground, drone, buoyTrack hand-off targets
- False-alarm restraintCore decision feature
Portfolio deck, slide 91: Proven in a simulatorPortfolio deck, slide 92: The simulator, running
What has to go right
Commercial scale hinges on navigating slow, multi-year defence procurement cycles and securing mandatory military security clearances. Management must resolve background intellectual property ownership, as past delivered revenue of ₹23.01L resided in a separate legal entity with IP owned by MCEME. DeepGrid must also successfully transition its early FPGA platform to qualified production silicon without procurement delays disrupting lumpy defense order intake.
Key dependency Defence procurement cycle and security clearances.
Further reading and viewing
Watch
The Defence D-HUMR in the narrated walkthroughOne silicon. The full story. From 26:32Film: Perimeter threat fusionRadar, thermal and camera tracks correlated at the fence · 0:49Read
The D100 Drone SoC on the SoC2 dieTechnology: six domains, one tapeoutInvestment memorandum: Strategy & economicsDefence-first, make the secure-compute chiplet the revenue engineFinancial Model v3 (Sept 2026), workbookSheet: Revenue BuildInformation Memorandum v2 (Aug 2026), page 209 B . C O M P E T I T I V E P O S I T I O N I N GAsk DeepGrid about the Defence D-HUMRAny question, answered from the same documentsRelated products
Prices, volumes, revenues and margins are management projections. The text is written from the linked documents, and every figure in it appears in the source it came from.