Rethinking Surveillance

For decades, surveillance has been treated as a reactive measure—responding after an incident occurs. The challenge, however, is that this approach often leaves gaps, delays, and inefficiencies. Soranet was designed to turn that model on its head. Instead of waiting for problems to unfold, it creates a proactive security ecosystem that combines aerial, ground, and digital intelligence into a single operational framework.

The outcome is not just more coverage—it’s faster, smarter decision-making.

The Core of the System: Integrating Multiple Layers of Intelligence

Soranet doesn’t rely on one source of truth. It merges different streams of data to create a unified operational picture:

• Live drone video feeds for real-time aerial monitoring

• Ground-based CCTV networks for persistent coverage

• ANPR (Automatic Number Plate Recognition) for vehicle identification

• Geospatial mapping to contextualize movement and boundaries

• IoT-powered access control systems, such as automated gates

All of this intelligence is orchestrated through a custom Android application and a responsive dashboard that brings decision-making tools directly into the hands of operators.

At the center is a fleet of custom-built drones. Unlike traditional manual operations, these drones are designed for autonomy: they patrol predefined routes, respond instantly to alerts, and adapt to geofenced rules. When a safety breach or intrusion is detected, the system can reassign drones in real time—making surveillance both flexible and resilient.

Takeaway: By merging multiple data sources into one operational dashboard, Soranet ensures no single point of failure defines security outcomes.

Intelligence on the Edge: From Detection to Decision

Building Soranet wasn’t just about deploying cameras in the sky. The true challenge was making those cameras smart enough to interpret what they were seeing. The engineering team developed computer vision models tailored to critical security use cases, such as:

• Vehicle detection and license plate recognition

• Intruder identification and posture analysis (e.g., distinguishing between an employee and a trespasser)

• Safety monitoring, including smoke, fire, or unauthorized access

• Movement cross-checks against pre-defined schedules and permissions

These models don’t just capture data—they analyze it in real time. Alerts are pushed instantly to operators, who can trigger automated actions such as raising alarms, opening or closing gates, or dispatching drones for closer inspection.

Takeaway: By embedding AI directly into the operational loop, Soranet shifts security from passive observation to active response.

Behind the Scenes: Solving Complex Systems Problems

Delivering such a system required tackling technical hurdles across software, firmware, AI, and hardware. The architecture included:

• Backend: Python, C++, and ROS (Robot Operating System)

• Frontend: Android (Kotlin-based dashboard and controls)

• Computer Vision: TensorFlow and custom OpenCV pipelines for classification

• Drone Firmware: APIs for remote commands and telemetry

• Infrastructure: Primarily on-premises, with pathways for cloud deployment

One of the hardest problems was bandwidth. Drones often operate in environments where connectivity is weak or unstable. To address this, the team implemented buffering, compressed models through quantization, and deployed inference directly on drone hardware.

Takeaway: Soranet’s design shows that true innovation in AI systems isn’t just about the algorithm—it’s about ensuring those algorithms work in imperfect, real-world conditions.

From Pilot to Deployment: Real-World Outcomes

Soranet is not a theoretical model—it’s already operational. In partnership with Argent Bright Security, deployments have been rolled out across UK parks and warehouses. It has been showcased at national conferences, recognized as an innovation in automated surveillance, and importantly, built with GDPR compliance from the ground up.

The results speak clearly:

• Surveillance coverage expanded by more than 300% in pilot areas

• False positives reduced thanks to sensor fusion and AI-based filtering

• Response times to incidents significantly shortened

• Active integration achieved with local law enforcement in certain deployments

Takeaway: Success in AI-driven surveillance is measured not only in innovation but in tangible improvements to security outcomes.

Lessons Beyond Surveillance

What makes Soranet notable isn’t just the drones or the AI—it’s the way the problem was framed. The project succeeded because the team combined systems thinking, rapid prototyping, and continuous feedback loops.

That mindset—problem-first, system-driven, execution-focused—is exactly what any ambitious technology project demands. Whether it’s automating workflows in business, integrating AI into government systems, or designing resilient industrial platforms, the principle holds: success starts with understanding what truly needs to be solved.

Takeaway: The approach behind Soranet is as valuable as the technology itself.

The Road Ahead: Expanding Capabilities

The story of Soranet is still being written. The next chapter includes:

• Integrating with military-grade drone systems for defense applications

• Expanding surveillance to industrial estates and critical infrastructure

• International pilots, including early-stage discussions with an African government for port security

• Scaling the system for cloud-based deployments to meet the needs of larger enterprises and government contracts

These steps signal that Soranet is not a one-off project but a platform capable of evolving to meet diverse security challenges.

Final Reflections

While our consultancy didn’t build Soranet, one of our co-founders was part of its original engineering team. That experience—solving problems at the intersection of AI, hardware, and real-world operations—shapes the DNA of our firm today.

We carry forward the same philosophy: complex problems require multidisciplinary solutions. Innovation isn’t just about writing code; it’s about aligning technology with strategy, compliance, and outcomes that matter.

For organizations navigating ambitious, technically demanding projects, that perspective makes the difference between a system that works in theory and one that succeeds in practice.