4D Radar Pod
The 4D Radar Pod is a sealed automotive radar module delivering range, velocity, azimuth and elevation at ₹25,000 per unit, sold into the aftermarket and integrator channel. Its commercial rationale rests on a single engineering fact: the R100 radar block is designed into the SoC2 die, so the one tapeout that carries the portfolio carries it too.
- Listed price
- ₹25,000
- FY2032 revenue
- ₹7.5 Cr
- Share of plan
- 0.7%
- Gross margin
- 50%
- First revenue
- FY2027
- Volume plan
- 3,000 @ FY32
From input to outcome
Proposed product architecture. The scene illustrates its application, not a shipped kit.
The radar illuminates the scene and reads the returns
The R100 block on the die turns returns into range, velocity and elevation
Tracks are formed and held across frames, including through rain and dust
Tracks fuse with the camera object list to confirm or reject detections
Drag to rotate. Conceptual layout, not a mask.
On the die
The case in brief
- By delivering direct physical measurement of closing speed through rain, dust, and glare, the pod provides the defensible inputs required for automated braking decisions where camera-only setups degrade.
- Projected at ₹7.50 Cr by FY2032 (3,000 units, 0.7% of group revenue), the pod functions primarily to complete the surround-sensing architecture for flagship compute systems rather than as an independent volume driver.
What it is used for
Fleet radar upgrade on camera-only fitments
- Buyer
- Fleets and integrators adding radar to camera-only fitments
- The job
- Camera-only ADAS systems cannot measure closing speed directly and degrade in rain, dust and glare, conditions common in Indian commercial-vehicle operations, leaving braking decisions based on estimated rather than measured data.
- What DeepGrid supplies
- The pod mounts forward or on a corner and feeds range, velocity and elevation tracks to the compute box; the R100 block on the SoC2 die fuses those tracks with the camera object list so that closing speed, the quantity a braking decision actually turns on, is measured, not estimated.
Adverse-weather surround sensing in integrated truck kits
- Buyer
- OEM cab and body engineers and Tier-1 integrators
- The job
- Full-surround ADAS kits for commercial vehicles require a sensor channel that continues to function when cameras degrade; front and rear radar coverage must be fused with multiple camera streams within a single frame budget.
- What DeepGrid supplies
- The pod provides front and rear 4D 77 GHz radar inputs that the SoC2 fuses with seven RGB and two thermal cameras in 8.6 ms of the 33.3 ms frame, supplying animal and cyclist detection in adverse conditions as part of a complete surround-coverage architecture.
Defensible braking-decision sensor for regulated applications
- Buyer
- Fleet safety officers and integrators where braking decisions must be defensible
- The job
- Where a collision event is reviewed after the fact, a closing-speed figure derived from camera estimation may not withstand scrutiny; operators and integrators need a measured value from a dedicated sensor.
- What DeepGrid supplies
- The radar measures closing speed directly, a quantity a camera estimates and a radar measures, and holds tracks across frames including through rain and dust, providing a recorded, sensor-derived input that supports post-event review.
The story
A sealed, all-weather radar pod built to feed the central compute box.
The 4D Radar Pod is a weatherproof unit mounted forward or on vehicle corners to provide dependable spatial data under environmental extremes. Priced at ₹25,000, it packages an imaging front end with DeepGrid's native radar processing inside an automotive-grade housing. It uses the radar block designed into the SoC2 die, so it needs no tapeout of its own.
- ₹25,000Unit price
- SensorsSegment
- OutdoorOperating domain
- FY2027First revenue
- 4D imaging radar front endComponent
- R100 radar DSP processingComponent
Hardware-level tracking directly measures closing speed in adverse operating conditions.
The pod illuminates the scene and processes raw returns through the on-die R100 block into range, velocity, and elevation. Persistent tracks are maintained across frames through dense dust, heavy rain, and harsh glare where vision sensors fail. These tracks fuse with the camera object list to supply the definitive closing speed required for emergency braking actions.
- Range, velocity, elevationCalculated outputs
- Measured closing speedDecisive metric
- Rain, dust, glare resilienceOperating advantage
- Front-end supply qualificationKey dependency
Fleets and integrators upgrading camera-only fitments for defensible safety records.
Commercial fleet operators and retrofit integrators purchase the pod to resolve critical camera blind spots created by weather, road spray, and dust. A measured radar closing speed withstands post-incident scrutiny where camera-estimated distances cannot. The offering addresses a $2.75B global market pool, requiring negligible market capture to fulfill its production plan.
- Fleets and integratorsTarget buyer
- Aftermarket and integrator supplyChannel
- Camera degradation in weatherAdoption trigger
- $2.75B global (2025)Demand pool
- 0.03% at FY2032Management capture projection
A high-margin sensor complement designed to unlock complete compute-suite deployments.
Generating a projected ₹7.50 Cr at 3,000 units by FY2032, the radar pod accounts for 0.7% of group revenue at a 50% gross margin. It exists because the flagship compute platform requires surround sensing, and the processing block is already paid for on the SoC2 die. The commercial business carries bounded liability by delivering raw sensor inputs directly into integrator-owned braking safety cases.
- 3,000 unitsFY2032 volume projection
- ₹7.50 CrFY2032 revenue projection
- 0.7%Group revenue share
- 50%Gross margin projection
- No incremental silicon costPortfolio role
- ₹0.07 CrFY2730 units
- ₹0.3 CrFY28120 units
- ₹0.88 CrFY29350 units
- ₹2 CrFY30800 units
- ₹4 CrFY311,600 units
- ₹7.5 CrFY323,000 units
What has to go right
Radar front-end supply and antenna qualification are the stated dependencies; the DSP processing is on DeepGrid's own silicon and is not at risk. Verification of the radar block on the ASIC must complete before the product can ship, making that milestone a gating item for FY2027 first revenue. Volume ramps from 30 units in FY2027 to 3,000 in FY2032, so the business depends on integrators and fleets committing to radar upgrades at scale rather than treating the pod as a one-off addition. Because the pod is sold as a component into integrators' safety cases rather than as a complete system, DeepGrid does not own homologation or the braking safety case, which bounds both the liability and the addressable scope.
Key dependency Radar block verification on the ASIC.
Further reading and viewing
Read
The R100 Radar SoC on the SoC2 dieTechnology: six domains, one tapeoutInvestment memorandum: Use casesWhat Deepgrid sells, and what it does notInformation Memorandum v2 (Aug 2026), page 10Our own SoC fuses 7 RGB + 2 thermal + front/rear 4D radar in 8.6 ms of the 33.3 ms frame (Information Memorandum (June 2026), PDFPage 11Information Memorandum (June 2026), PDFPage 10Business Plan v2, workbookSheet: MarketFinancial Model v3 (Sept 2026), workbookSheet: Demand & TAMBusiness Plan v2, workbookSheet: AssumptionsAsk DeepGrid about the 4D Radar 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.