Thermal Camera Pod
The Thermal Camera Pod is a bolt-on long-wave infrared sensor module with an integrated lens heater and weatherised housing that feeds into the SoC2 compute box. Serving as the perception layer for night and monsoon operation, management projects the product to generate ₹16.0 Cr in FY2032 revenue at a 50% gross margin as a capability line closing the condition where visible cameras are weakest.
- Listed price
- ₹50,000
- FY2032 revenue
- ₹16 Cr
- Share of plan
- 1.4%
- Gross margin
- 50%
- First revenue
- FY2027
- Volume plan
- 3,200 @ FY32
From input to outcome
Proposed product architecture. The scene illustrates its application, not a shipped kit.
The LWIR sensor reads emitted heat rather than reflected light
SoC2 runs detection on the thermal image, where a person is bright against cold road
Classification separates people and animals from roadside clutter
Detections join the same object list the cameras feed
Drag to rotate. Conceptual layout, not a mask.
On the die
The case in brief
- Management projects the Thermal Camera Pod to generate ₹16.0 Cr in FY2032 revenue across 3,200 units, representing a 1.4% group revenue share at a 50% gross margin.
- Positioned as an aftermarket capability attach priced at ₹50,000, the pod feeds the SoC2 compute box to sustain detection when optical cameras fail in night, monsoon, dust, or smoke.
- The product functions as a portfolio defense rather than a core margin driver, carrying its primary operational risk in imported thermal sensor supply.
What it is used for
Aftermarket Fleet Night and Monsoon Retrofit
- Buyer
- Fleets upgrading existing vehicles
- The job
- Commercial vehicle fleets face sharp perception degradation when operating visible-light cameras through night, monsoon rain, glare, dust, and smoke.
- What DeepGrid supplies
- DeepGrid supplies a bolt-on long-wave infrared sensor module with a weatherised housing, lens heater, and harness that connects to the SoC2 compute box to detect emitted heat from people and animals.
Smart Truck ADAS Tower Integration
- Buyer
- OEM cab and body engineer, Tier-1 HMI owner, and fleet safety officer
- The job
- Heavy commercial vehicles require continuous blind-spot and moving-off monitoring under AIS-186 and AIS-187 without high nuisance alert rates when visibility conditions collapse.
- What DeepGrid supplies
- DeepGrid integrates thermal sensing into retrofit mirror towers as one of eleven fused perception channels feeding directly into the SoC2 object list.
The story
A bolt-on long-wave infrared sensor pod that feeds existing compute boxes to resolve zero-visibility conditions.
The Thermal Camera Pod packages a long-wave infrared sensor, a dedicated lens, and an integrated heater inside a weatherised outdoor housing. Designed as an aftermarket capability upgrade priced at ₹50,000, the module connects directly via harness into the SoC2 compute box. It closes severe operating gaps where primary optical sensors degrade, with first revenue scheduled for FY2027.
- ₹50,000Unit price
- SensorsSegment
- OutdoorOperating domain
- FY2027First revenue
- LWIR sensor and lensHardware component
- Weatherised housing, lens heaterHardware component
Emitted heat detection allows SoC2 to classify vulnerable road users and merge detections into the primary object list.
The LWIR sensor reads emitted heat rather than reflected light, making pedestrians and animals stand out against cold road surfaces through unlit conditions, glare, dust, and smoke. The SoC2 processor executes detection on the thermal stream, classifies obstacles away from roadside clutter, and routes them into the common object list alongside optical feeds. Thermal sensor supply is imported, representing the primary bill-of-materials risk in the sensor group.
- Emitted heat detectionSensing principle
- SoC2 thermal detectionCompute processing
- People and animal classificationClassification stage
- Common object listData fusion
- Imported thermal sensor supplyPrimary supply dependency
Commercial vehicle fleets attach thermal pods as aftermarket upgrades for night and monsoon haulage.
Purchasing is driven by fleets running high-risk night and monsoon operations where camera-only perception degrades sharply. DeepGrid distributes the pod through aftermarket channels, offering it as an add-on attach alongside AD0 and AD2 retrofits at the same fitment. Management assumes 0.10% capture of the $1.68B global automotive thermal camera market by FY2032.
- Fleets upgrading existing vehiclesTarget buyer
- Attach alongside AD0/AD2Sales model
- AftermarketDistribution channel
- Night and monsoon operationAdoption trigger
- $1.68B global (2025)Demand pool
- 0.10% (management projection)FY2032 pool capture
A 50% gross margin capability line designed to harden the broader system rather than drive group margins.
Management projects the Thermal Camera Pod to generate ₹16.0 Cr in FY2032 from 3,200 annual units, representing 1.4% of total group turnover. The unit operates at a 50% gross margin, below corporate average, reflecting substantial physical sensor component costs. Its strategic role is to eliminate edge-case failure modes for higher-tier systems, with volume expansion governed by imported sensor supply.
- ₹16.0 Cr (management projection)FY2032 revenue
- 3,200 units (management projection)FY2032 volume
- 1.4% (management projection)Group revenue share
- 50% (management projection)Gross margin
- Capability line, not profitPortfolio role
- Imported thermal sensor supplyCritical dependency
- ₹0.2 CrFY2740 units
- ₹0.75 CrFY28150 units
- ₹2 CrFY29400 units
- ₹4.5 CrFY30900 units
- ₹9 CrFY311,800 units
- ₹16 CrFY323,200 units
What has to go right
DeepGrid depends on imported thermal sensor supply, which represents the highest bill-of-materials risk in the sensor group. The engineering team must maintain a robust sensor sourcing and calibration pipeline to reliably fuse thermal detections with visible camera feeds. In commercial vehicle deployments, DeepGrid operates strictly as a bounded camera-perception subsystem, requiring measured false-information testing and qualification through independent testing agencies such as ARAI or ICAT.
Key dependency LWIR sensor sourcing and calibration pipeline.
Further reading and viewing
Read
The T100 Transformer NPU on the SoC2 dieTechnology: six domains, one tapeoutInvestment memorandum: Use casesWhat Deepgrid sells, and what it does notCommercial use case UC-03Blind-spot and moving-off informationPortfolio deck, slide 26Each sensor's jobAsk DeepGrid about the Thermal Camera PodAny 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.