FAQs

Stellarview is an Indian edge AI company that designs and manufactures video analytics products for traffic enforcement, smart infrastructure, and access control. We are a subsidiary of Adagrad AI Pvt Ltd, headquartered in Pune. Our products include ANPR cameras, speed detection systems, incident detection systems, mobile enforcement platforms, and access control devices. All hardware is designed, engineered, and assembled in India.

We serve five primary verticals: highways and expressways (enforcement, incident detection, traffic analytics), smart cities (urban surveillance and automated enforcement), toll plazas (vehicle classification and lane ANPR), residential and gated communities (access control), and commercial and industrial properties (enterprise access control). Our products are deployed across 25,000+ camera installations and monitor over 1 million vehicles daily.

We work with system integrators, infrastructure companies, and technology partners on a project-by-project basis. If you are a system integrator, toll operator, concessionaire, or smart city solution provider looking to integrate Stellarview's camera or analytics platform into your deployment, write to us at sales@stellarview.ai with your project details and we will set up a technical discussion.

Fill out the contact form on our website or email sales@stellarview.ai with your use case, location, and requirements. Our team will schedule a demo (remote or on-site) within 48 hours. For access control (GateGuard), we also offer free site surveys to assess your gate setup before deployment.

We provide both on-site and remote support. For GateGuard deployments, we offer Annual Maintenance Contracts (AMC) that cover hardware servicing, software updates, and on-site visits. For highway and enforcement deployments, our team provides remote monitoring, firmware updates, and field support through our operations office in Pune. Our deployment and customer support team makes up 20% of our workforce to ensure consistent service quality.

All data is stored on India-based servers. Evidence images, vehicle logs, and incident records are retained for 1 year by default. For GateGuard Connect cloud storage, retention duration depends on the subscription tier. On-premises deployment is available for defense, government, and enterprise clients who require data to stay within their own infrastructure.

Yes. Our systems are STQC certified for security assurance and support encrypted data transmission. The EdgeTrack dashboard and GateGuard Connect both use role-based access control, ensuring that each user sees only what is relevant to their role. Full audit logging tracks all system and user activity. For on-premises deployments, no data leaves the client's network.

All products are designed, engineered, and assembled in India. Our hardware (including the Recon Global Shutter Camera) is manufactured in-house and carries the Make in India certification. Our core software stack is built entirely in C/C++ with GStreamer for high-performance inference, running on NVIDIA edge processors.

At the company level, we hold ISO 9001 (quality management), ISO 27001 (information security), ISO 14001 (environmental management), and ISO 45001 (occupational health and safety). Our products carry NATRAX certification (speed detection up to 230 kmph), STQC certification, BIS certification, and full compliance with the NHAI 2023 VIDES circular. We are also NASSCOM Emerge 50 winners and Maruti Suzuki Accelerator Cohort 9 winners.

Visit the Careers page on our website to view current openings across engineering, sales, deployment, and data operations. We are a team of 40+ people across technical development, management, sales and marketing, data labelling, and deployment support. If you do not see a role that fits but believe you can contribute, send your resume to the email listed on the Careers page with a short note about what you would like to work on.

A global shutter captures the entire image frame at once, unlike a rolling shutter that scans line by line. This eliminates motion blur on fast-moving vehicles. At 230 kmph, a rolling shutter produces distorted plates. A global shutter produces a clean, readable image every time.

HSRP, non-HSRP, Bharat Series, Green (EV), Yellow (commercial), White (private), hand-painted, and non-standard plates. It supports single-line, double-line, and triple-line formats. Trained on 10M+ Indian plates.

Plate (ANPR) recognition accuracy of 96% for readable number plates, day and night. Vehicle detection rate above 99.9%. These figures refer to different metrics and should not be combined into a single headline number.

NATRAX certified for radar-based speed detection up to 230 kmph and vision-based detection up to 180 kmph.

Yes. The camera has built-in IR with a range of up to 120 meters and NightColor technology for colour imaging in low-light conditions. Plate and violation detection accuracy is maintained after dark.

Overspeeding (vision and radar), no helmet (rider and pillion), no seatbelt (driver and passenger), triple riding, wrong direction, mobile phone use, lane discipline violation, and vehicles not allowed on the stretch.

AI Search is an optional add-on that uses a Vision Language Model to let operators query camera feeds in natural language. For example, "man in orange turban" or "truck with Ganesh written on it." Results are returned in seconds.

STQC certified, BIS certified, NATRAX certified (230 kmph), NHAI 2023 VIDES policy compliant, ONVIF supported (Profile S, T, M, G), MLFF and GNSS ready, and Make in India.

VIDS is an AI-powered system that detects traffic incidents in real time using video analytics. It is built to NHAI's 2023 policy and designed to trigger immediate alerts to control rooms when dangerous situations occur on the road.

Fallen objects on the road, accidents, fire and smoke, poor visibility (fog, heavy rain), stalled vehicles, slow-moving vehicles, wrong-direction driving, pedestrian and animal crossings, and camera tampering.

Incidents are detected within seconds of occurring. Alerts with evidence images and video clips are pushed to the EdgeTrack dashboard, and optionally via SMS and email to control room operators.

Yes. All AI processing happens on the Local Processing Unit at the gantry site. Incidents are detected, logged, and stored locally. Data syncs to the central server when connectivity is available.

Each VIDS LPU supports up to 3 cameras, enabling multi-lane coverage from a single processing unit.

Yes. The AI model uses duration, location (lane vs shoulder), and movement patterns to distinguish between stalled vehicles requiring alerts and legitimately stopped vehicles on the shoulder.

Yes. Each detected incident produces both an evidence image (snapshot) and a short evidence video clip for detailed analysis. Both are logged with timestamp, location, and incident type.

Yes. VIDS is built to the NHAI 2023 VIDES circular, the 2021 circular, and the 2016 circular. It meets all functional and technical requirements specified for video-based incident detection on national highways.

VSDS is a speed detection and enforcement system that combines a camera with a 4D Doppler radar to measure vehicle speeds and detect traffic violations. It generates evidence images with speed readings for automated challan generation.

The 4D Doppler radar measures the speed of each vehicle across all lanes. It can track up to 120 vehicles simultaneously. The camera captures evidence images tied to each speed reading.

Yes. VSDS also detects no helmet (rider and pillion), no seatbelt (driver and passenger), triple riding, wrong lane, wrong direction, and vehicles not allowed on the stretch.

Yes. Spot speed is measured at the gantry location. Section speed is calculated using the travel time between two VIDES gantries (typically 10 km apart), as specified in the NHAI 2023 circular.

Yes. The system supports separate speed limits per vehicle class. For example, commercial vehicles can be enforced at 80 kmph while cars are enforced at 120 kmph on the same stretch.

Yes. The system uses built-in IR and Pulse IR technology for clear evidence capture in complete darkness. Night-time seatbelt detection for both driver and co-driver achieves 96% accuracy.

Each violation produces a timestamped, geo-tagged evidence image showing the vehicle, plate (with OCR), speed reading, violation type, and lane position. This evidence is designed to stand scrutiny in court.

A Recon Global Shutter camera, a 4D Doppler radar, a Local Processing Unit (LPU), and optionally a speed display board. All components mount on a standard gantry.

ATCC is an AI-based system that automatically counts and classifies every vehicle passing under a gantry. It provides real-time traffic volume data broken down by vehicle type, direction, and time.

18+ classes including 2-wheelers, 3-wheelers, 4-wheelers (cars, jeeps, vans), LCVs, mini LCVs, buses, trucks (2-axle through 6+ axle), tractors, heavy construction machinery, and goods autos.

Vehicle counting accuracy of 99.9%. Classification accuracy of 97% or higher across all vehicle types in both day and night conditions.

Yes. The system captures evidence images for every counted vehicle in both day and night conditions using the camera's built-in IR.

Vehicle count by class, direction, and time interval. Evidence images per vehicle. Hourly, daily, and custom reports. Data is available on the EdgeTrack dashboard and exportable to Excel.

Yes. ATCC is widely used at toll plazas to verify vehicle classification against the toll category charged. It supports axle-based classification for accurate multi-axle truck categorization.

Yes. ATCC meets the NHAI 2023 VIDES circular requirements for automatic traffic counting and classification. Vehicle-wise data is stored for comparison with NPCI FASTag transaction records.

Yes. The NHAI 2023 circular specifies co-locating ATCC, VSDS, VIDS, and ANPR on a single VIDES gantry. This reduces hardware costs and simplifies maintenance.

VIDES is the complete, integrated system that combines VSDS (speed detection), VIDS (incident detection), ATCC (traffic counting), and ANPR (plate recognition) on a single gantry. It is the system specified in NHAI's 2023 circular.

The 2023 circular mandates co-locating all four functionalities on one gantry every 10 km. This reduces infrastructure costs, enables shared sensor use, simplifies maintenance, and provides a unified data pipeline.

It detects overspeeding and violations (VSDS), identifies incidents like accidents and fallen objects (VIDS), counts and classifies every vehicle (ATCC), and reads every number plate (ANPR). All from one installation.

All gantry data flows to the EdgeTrack centralized dashboard via fibre or internet. Operators see violations, incidents, traffic data, and device health in one unified view.

Violations detected at the gantry are sent to EdgeTrack for officer validation. Approved violations are pushed to NIC for automated e-challan generation and delivery to the vehicle owner.

NHAI specifies VIDES gantries every 10 km on both sides of the highway. Each gantry covers all lanes including shoulder portions.

Yes. The NHAI circular requires the ability to continuously update cameras with improved AI models and newer incident types remotely, without physical changes to the hardware setup.

VIDES deployments require NATRAX certification for speed detection, STQC certification for security, NIC integration for e-challan, and compliance with the full NHAI 2023 technical specification. Stellarview meets all of these.

MTES is an AI-powered system that transforms patrol vehicles into smart enforcement units. A roof-mounted camera, 4D Doppler radar, and on-board processor detect violations in real time while the vehicle is parked or moving.

MTES requires no permanent infrastructure. It can be deployed at accident-prone zones, blackspots, or shifting locations. It brings enforcement to areas where building a gantry is not feasible or cost-effective.

Overspeeding, no helmet (rider and pillion), no seatbelt (driver and passenger), triple riding, non-standard number plates, and document violations cross-referenced with Vahan and Sarathi databases.

The 4D Doppler radar tracks up to 120 vehicles simultaneously across all lanes, with accurate speed measurement up to 200+ kmph.

Yes. The 9MP camera has IR capabilities for full night operations. Evidence images including ANPR are captured in complete darkness.

Two-level verification. Level 1: the on-site enforcement officer reviews detections on the local dashboard. Level 2: an RTO officer reviews and approves on the cloud dashboard. Approved violations are pushed to NIC for e-challan.

One click. The officer starts a session, GPS auto-logs the location, and the system is active. No manual calibration or complex setup required.

69 interceptor vehicles deployed statewide in Maharashtra for the Motor Vehicles Department and Maharashtra Police. It is the largest mobile traffic enforcement deployment in India.

GateGuard is an edge AI access control system that reads vehicle number plates and controls gate barriers automatically. Registered vehicles get a green signal and the barrier opens. Unregistered vehicles get a red signal and the guard is alerted.

Above 98% for 4-wheelers and above 92% for 2-wheelers. It runs inference at 25 FPS, giving multiple chances to read each plate accurately.

Yes. GateGuard is specifically trained on Indian bikes, scooters, and auto-rickshaws. No other ANPR engine in India works on 2-wheelers at 25 FPS.

No. ANPR reads the existing number plate. There is no tag, sticker, or per-vehicle cost. The investment is in the hardware and the optional GateGuard Connect subscription.

All AI processing runs on the edge device locally. Access decisions, barrier control, and evidence logging continue without internet. Data syncs when connectivity returns.

GateGuard's ANPR engine is trained on 10M+ Indian plates including hand-painted, non-standard, partially hidden, and colour plates (commercial, EV, defense, diplomatic). It is colour-independent and font-independent.

Through the GateGuard Connect web portal or the guard mobile app. Vehicles can be added individually or via bulk CSV/Excel upload. Remote editing is available from any browser.

GateGuard Connect is the cloud platform that lets property managers oversee one gate or hundreds from a single screen. It adds remote management, automated reporting, and analytics on top of the GateGuard edge device.

Add, edit, or delete whitelist/blacklist entries. View live and historical device status (online/offline). Pre-register expected vehicles. Download evidence images. View analytics and TAT data.

Blacklisted vehicle entry, vehicle overstay, specific vehicle alerts (e.g., VIP arrival), and device downtime notifications. Alerts are delivered via email.

Daily summaries including entry/exit counts, 2W vs 4W breakdown, visitor logs, uptime data, and TAT analytics. Delivered automatically via email to designated recipients.

Essential (basic remote access and whitelist editing), Professional (adds automated reporting, email alerts, TAT, and cloud storage), and Premium (adds custom reporting, custom fields, evidence camera support, parking control, and analytics dashboard).

Yes. GateGuard Connect supports multi-site management. All gates across all properties appear on one dashboard with centralized whitelist, reporting, and device monitoring.

Yes. SDK includes Whitelist API (add, edit, delete vehicles programmatically) and Events API (receive real-time entry/exit events via webhook).

Yes. For defense, government, and enterprise clients requiring data to stay on-site, GateGuard supports on-premises deployment alongside the cloud Connect platform.

EdgeTrack is the centralized monitoring and enforcement platform that connects all Stellarview edge devices. It provides a unified dashboard for violations, incidents, traffic data, device health, and e-challan generation.

ANPR, VSDS, VIDS, ATCC, and VIDES. All data from all connected edge devices flows into one interface, regardless of whether the device is on a highway gantry, at a toll plaza, or in a city.

View real-time violation and incident feeds. Filter by location, date, vehicle type, or violation type. Review evidence images. Approve violations for challan generation. Monitor device uptime and health. Generate and download reports.

Yes. EdgeTrack connects directly with the National Informatics Centre. Validated violations are pushed for automated e-challan issuance, reducing manual paperwork and ensuring legal compliance.

Dashboard alerts, SMS notifications, and email alerts. Alerts can be configured for incidents (accident, fire, fallen object), device downtime, specific vehicle detections, and system issues.

Highway patrol officers, NHAI PIUs, concessionaires, toll operators, control room staff, city traffic departments, and system integrators. Role-based access ensures each user sees only what is relevant to them.

Yes. The dashboard includes a GIS corridor view showing device locations, live status, and incident markers on a map. This is particularly useful for highway deployments spanning long corridors.

EdgeTrack supports both. Cloud deployment is standard. On-premises deployment is available for government and defense clients with data residency requirements.

AI Search is an advanced search feature that uses a Vision Language Model (VLM) to let operators query camera feeds in natural language. Instead of scrubbing hours of footage, you type a description and get matching results in seconds.

Text on vehicles (e.g., "Ganesh" written in Marathi), brand logos (e.g., "Puma"), appearance attributes (e.g., "man with red cap," "yellow face emoji sticker"), vehicle descriptions (e.g., "damaged white truck"), and partial plate matches.

English, Hindi, and Marathi. The system recognizes text in all three languages on vehicles, signs, and other surfaces.

Results are returned in real time with minimal latency. The search results update dynamically as new data is added from the camera feed.

AI Search runs on the Recon Global Shutter camera's edge processor (157 TOPS, 1024 CUDA cores). No server-side processing is needed for the search itself, though results are viewable on the EdgeTrack dashboard.

Law enforcement investigations (finding suspect vehicles or people by description), highway monitoring (finding specific vehicles across corridors), and forensic analysis (searching recorded footage for specific events or objects).

AI Search is an optional add-on to the Recon Global Shutter camera. It can be enabled on any deployed camera without hardware changes.

Traditional ANPR search is limited to plate numbers. AI Search goes beyond the plate. You can search by vehicle appearance, text, logos, human attributes, or any visual detail the camera has captured. It answers "what did you see?" not just "which plate passed."