PERCEPTA
Edge Visual Intelligence
for the Real World
A multi-purpose computer vision platform that counts, measures, and analyses without identifying individuals. From traffic analytics to security monitoring to custom integrations, turn any camera into actionable insights.
Solutions
One platform. Multiple applications. Traffic is our flagship demo, but the same computer vision technology extends to footfall, security, facilities, events, and beyond. Models, alert rules, and features are configurable per tenant, and we build new ones to order.
Traffic & Mobility
Vehicle counting, classification, and speed monitoring for roads, junctions, and car parks. Evidence capture for traffic calming initiatives.
- ›Vehicle counts by type
- ›Speed zone monitoring
- ›Junction flow analysis
- ›Traffic calming evidence
Footfall & Public Spaces
Anonymous pedestrian counting for high streets, town centres, parks, and trails. Understand peak times and flow patterns without identifying anyone.
- ›High street footfall
- ›Park & trail usage
- ›Public space planning
- ›Active travel monitoring
Facilities & Sites
Entrance flow monitoring for campuses, buildings, and managed sites. Aggregate occupancy trends and utilisation insights.
- ›Campus entrance counts
- ›Building occupancy trends
- ›Car park utilisation
- ›Queue length monitoring
Events & Venues
Crowd flow patterns for festivals, venues, and temporary events. Aggregate data for operational planning and safety.
- ›Entry/exit flow rates
- ›Peak time identification
- ›Capacity trend monitoring
- ›Post-event reporting
Security & CCTV Analytics
Turn existing CCTV into an intelligent monitoring layer. People detection, loiter time analysis, zone-based alerting, and near real-time notifications.
- ›People detection & counting
- ›Loiter time monitoring
- ›Zone activity alerts
- ›No-activity detection
Have a different use case?
We are a small, agile team that builds fast. Custom alert rules, new detection models, bespoke dashboards, or full integrations into your existing workflows. Tell us what you need and we will get it done.
All solutions use the same privacy-first platform. No facial recognition. No personal identification.
Platform Capabilities
One platform, many applications. Privacy-first computer vision for traffic, footfall, security, and beyond.
Computer Vision
Advanced AI models detect and classify objects in real-time. Vehicles, pedestrians, cyclists, and more, all processed on-device.
Counting & Analytics
Accurate counts by type and direction. Identify peak times, flow patterns, and utilisation trends.
Speed Measurement
Calibrated speed zones for traffic applications. Automatic threshold monitoring and evidence capture.
Edge Processing
All AI runs on-device. No video leaves the site. Low bandwidth, offline-resilient operation.
Cloud Dashboard
Anonymised statistics sync to your secure portal. Real-time views, historical reports, CSV exports.
Near Real-Time Alerting
Built-in and custom alert rules per device and location. We can build alert types to match your specific operational needs.
Privacy by Design
No facial recognition. No personal identification. Only aggregate counts and flow statistics.
How It Works
From installation to insights in three simple steps. Privacy-first by design.
Deploy Edge Device
Install our compact Jetson-powered device at your site. Connects to any IP camera. Works indoors or outdoors.
$ percepta deploy --site="Main St" > Initialising device... > Camera connected ✓ > AI model loaded ✓ > Ready to monitor
AI Processes Locally
Computer vision runs entirely on-device. Detects, classifies, and counts in real-time. No video leaves the site.
> Frame 142857 processed > Objects detected: 5 > ├─ Vehicle (94.2%) > ├─ Vehicle (91.8%) > └─ Pedestrian (89.3%) > Counts updated ✓
Insights Sync to Cloud
Only anonymised statistics sync to your secure portal. View real-time dashboards, generate reports, export data.
> Syncing statistics... > ████████████████ 100% > Events uploaded: 847 > Dashboard updated ✓ > portal.percepta.io
Get Started
Ready to turn your cameras into insights? Let's talk.