SynaEYE and SynaHUB at a glance
With SynaEYE and SynaHUB, we are introducing two of our own European visual AI platforms. SynaEYE continuously analyses camera streams. SynaHUB answers targeted visual inspection questions using individual images. This means we are expanding our work as an integrator and operator to include our own technology for security, condition controls and operational processes.
You want to know early if someone is entering a restricted area. You may also need to regularly check whether a passage is clear or a gate is closed. Or you may want to investigate an incident after the fact without reviewing hours of video footage.
Such tasks begin with a specific question to the camera image. They differ in when the information is needed and which step should follow the result. This is exactly what we are focusing on SynaEYE and SynaHUB.
Why we are developing two platforms of our own
In our projects we encounter different requirements: A company needs continuous event detection across multiple cameras. Another wants to check a certain condition at fixed times. Still others need their own models or an application that combines multiple observations with operational rules.
With our own developments, we can tailor the analysis and its further development to such tasks. We combine our experience in integration and operation with the development of our own software.
The two platforms have different processes: SynaEYE accompanies ongoing events. SynaHUB organizes individual, planned or triggered test orders. Depending on the task, one platform is used or we combine both in a coordinated solution.
SynaEYE: Recognise events and check connections
SynaEYE is our platform for continuous AI video analytics. It combines event detection, object tracking, rules, alerting and forensic research. This allows current processes to be checked as well as events that have already been recorded to be specifically retrieved.
An example: An area on your company premises may not be entered outside of working hours. You define the relevant zone, the time window and the conditions for a report. SynaEYE evaluates the camera stream based on the analyses and rules set up. If a suitable event occurs, the responsible authority receives a note with image context.
Your employees can check the process and take the agreed next step. Depending on the connection, a control centre can be informed, a processing process can be started or an event can be transferred to an existing system.
The platform also helps with later research: time period, camera and visible object features narrow down the search. The information available depends on the analytics, retention, and permissions set up.
You can add your own models, object classes and rules for individual tasks. Image questions formulated linguistically can also support the specialised analyses. This allows us to tailor the detection to specific situations and processes in your company.
SynaHUB: An image question as a repeatable test task
SynaHUB is aimed at tasks in which an image needs to be evaluated at a specific point in time. You choose an image source, formulate a question and determine when the exam will take place.
For example:
- “Is the marked passage clear?”
- “Is the gate closed after hours?”
- “Are all the intended components in place?”
SynaHUB processes such tasks using vision-language models. These AI models evaluate images together with linguistically formulated questions. This creates a result that can be documented and passed on to your team or a connected system.
Test jobs can be triggered on a schedule, at intervals, on request or by an event. If necessary, rules can require a recheck or multiple confirmations before reporting. This makes it possible, for example, to distinguish whether an area is only recognised as blocked for a short time or repeatedly.
We use photographs from your company to check how reliably an image question can be answered. Viewing angles, lighting, occlusions and the selected model determine which tasks can be implemented sensibly.
From the image source to the next work step
The practical benefit arises when an analysis result is integrated into your process. To do this, we connect four building blocks:
Source → Analysis → Logic → Action
- Source: A suitable camera or a connected system provides the image material.
- Analysis: Suitable models recognise objects, movements or visible states.
- Logic: For example, rules take into account zone, time, duration and required confirmations.
- Action: The result is documented, handed over to a team or further processed in another system.
We determine together who needs what information and what should happen with it. This can include interfaces, notifications or a customized user interface. In this way, the analysis becomes part of a comprehensible operational process.
Combine continuous detection and targeted testing
Both approaches can work together for a delivery. One possible implementation: SynaEYE detects a defined event in the loading area. A test order is then started in SynaHUB via a suitable connection, which evaluates whether the intended area is free again.
The additional image inspection therefore takes place at a suitable time. We coordinate the trigger, test time, interface and follow-up action to suit your process. In a pilot, we use typical and difficult examples to check whether the combination delivers the required results.
Incorporate existing cameras sensibly
Suitable cameras and existing video systems can remain part of the solution. We check the technical connection as well as the viewing angle, resolution, lighting and frame rate.
The requirements arise from your task: A camera can show a zone violation well, while a different perspective is necessary to check a small component. That’s why we evaluate each image source based on the features that must actually be recognizable.
Based on this, we plan analysis, interfaces and operation. We add additional hardware where the task requires it.
European in-house development with control over operations
As a company based in Zurich, we at Synaedge combine software development, integration and operation. With SynaEYE and SynaHUB, we shape how our platforms develop and fit into our customers’ environments.
Both platforms can be operated with local computing resources and their own servers. Coordinated hybrid architectures are also possible, in which components work together in different locations.
Which form of operation is suitable depends on your locations, the required computing power and the protection needs of your data. That’s why we plan from the start which images and results will be processed, where they will be stored and who will be allowed to access them. We also include external model services and support access in this analysis.
The European development approach is complemented by concrete decisions about data paths, access and operation.
Vaidio remains part of our portfolio
With our own platforms, we expand the selection for your projects. Vaidio We continue to integrate and operate as an established enterprise platform.
We decide which technology to use based on functional requirements, existing systems and the operating model. Own platforms, partner technology and individual development can be combined appropriately. What matters to you is how well the entire solution fulfills your task.
What visual task do you want to solve?
With SynaEYE and SynaHUB, we combine continuous video analytics and targeted image checks with your company’s processes.
Would you like to recognise relevant events earlier, regularly check a visible condition or support manual control? Describe your situation to us. Together we examine the available images, determine the expected results and plan a manageable pilot.
