AI-Powered Video Surveillance Control System for a Mobile Security Equipment Provider
-
Industry
Security
-
Project type
Software development, AI integration
Softacom in Numbers
people from Softacom were working on the project
camera types were connected, detecting changes within ±1 second
months of work were saved via component reuse
reduction in false alerts
Description
A US-based provider of mobile surveillance equipment turned to Softacom to help build a smart and scalable system to control video surveillance operations and link footage directly to specific events.
The company delivers hardware that operates from an automated and self-regulating platform. They wanted to create a video surveillance control system that would link video pieces to events. They had made attempts to develop such a solution, created several components, but it was a long road to the end.
They needed a development partner who could help formalize their goals and deliver a multi-functional platform.
- No unified system to record, process and display camera events
- High costs of adapting to new camera models and sensors
- 3 partially developed modules, none integrated
- Need for 100% mobile operations as users in the field required real-time access
- No clear technical specification or documented requirements
Project team size
- 3 software developers
- 1 QA
- 1 PM/BA
Solutions
We started reviewing the client’s technical and product teams’ existing components and understanding what had already been built, what could be reused, and what still needed to be developed from scratch.
During this phase, we discovered that the requirements were incomplete. So we organized a series of planning sessions, where we formalized the system architecture. From there, we began building the system step by step:
Camera agents
They were designed to monitor surveillance cameras in a proactive mode, streaming live footage and detecting motion or environmental changes. When something unusual occurred, the agents would immediately transmit a signal and metadata to the server through a dedicated channel.
In the end, we integrated 20+ camera types that detect changes in ±1 sec.
Decision-making engine on the server side
It was made configurable so that the client could define logic: linking inputs from multiple cameras and sensors, interpreting 50+ complex event patterns, and notifying specific user groups. For example, if a motion sensor was triggered near a door and a camera detected movement at the same time, the system could automatically mark it as a security event and alert the appropriate team.
Content storage server
It archives each event with its associated video and photo files. It acts as a ventral media repository, allowing other parts of the system to request and review relevant footage.
Mobile application for Android and iOS
From their smartphones, users can monitor live events, escalate or reassign incidents, and view historical data. New mobile application delivers 100% mobile access to all surveillance events.
An intelligent decision agent trained to analyze video and image content
This is an AI-powered component that distinguishes between people, animals, and objects. It helps the system make smarter decisions. We used 10,000+ image samples to train the models.
For instance, if a motion sensor goes off and the camera detects a cat rather than a person, the AI tells the system to ignore it. It can also recognize vehicles, employees, and other predefined categories, reducing noise and false alarms. The false alerts were reduced by 54% during internal testing.
We completed the system and conducted thorough testing across all components. Currently, the client’s sales team is piloting the full solution with their own customers, gathering real-world feedback before a full-scale rollout.
Outcomes
- We delivered and tested a multi-component surveillance system.
- We enabled mobile-first monitoring that helps to make decisions on the go.
- Reduced false alarms by 54% thanks to AI filtering.
- Reused client assets, reducing time-to-market by 2 months and dev costs.
- Currently being piloted with end customers to validate product-market fit
“Softacom helped us transform our half-built idea into a fully operational platform ready for the market.”
– Technical Lead, Provider of Mobile Surveillance Equipment
