.webp){kind=logo}
icetana AI’s common hardware specification on a single server supports up to 200 cameras at 720p @ 15 FPS. For this scale, each server being connected via a 1 Gbps network interface is sufficient. If multiple servers are connected to a switch, the uplink facing the cameras should be 10 Gbps or higher.
For any other hardware requirements, please contact your Icetana AI representative for a detailed review.
icetana AI is optimised for H.265 (recommended for best performance and efficiency) and also supports H.264.
While MJPEG is supported, it is not recommended as it uses significantly more bandwidth and GPU resources, which may negatively affect system performance.
icetana AI is optimised for efficient processing but requires stable bandwidth for high-resolution video. Lower frame rates and resolutions can be configured to support low-bandwidth environments.
If your server reaches its capacity limits, such as CPU, GPU, or storage, icetana AI will begin dropping video frames to maintain performance. To help prevent this, our system includes built-in monitoring tools (using Grafana) that provide real-time insights into how each hardware resource is being utilised.
Your data remains securely stored on your servers, icetana AI cannot access your data without explicit permission. If you want to offload the PII responsibility, you can refer to T&C’s or EULA.
Yes, you can choose how long you’d like your data to be stored. We’ll work with you to configure retention settings that align with your operational and compliance needs.
No. When using icetana AI’s core Safety and Security product (without Facial Recognition), no personal or biometric data is stored. If you are using the Facial Recognition feature, icetana AI stores photographs of individuals listed in watchlists and event logs, in accordance with your configured settings and privacy policies.
Yes, icetana AI’s core Safety and Security product is GDPR compliant. It does not store any personally identifiable information. Additionally, video footage retention is fully configurable and can be aligned with your organisation’s specific security and data retention policies.
icetana AI’s Triage Agent allows security operators to define specific prompts or criteria they want to prioritise. The Triage Agent then analyses incoming Safety & Security events against these prompts. If an event matches the defined criteria, it is automatically escalated to the Livewall for immediate operator attention, helping reduce alert fatigue and focus on the most relevant incidents.
No, Forensic Quick Find currently supports tracking of people only. Vehicle tracking is not available at this time, but our development team continuously evaluates feature enhancements and their performance trade-offs based on customer needs.
Forensic Quick Find enables rapid investigation by allowing users to visually track individuals of interest across recorded footage. Once an event is generated by the Safety & Security module, you can:
This tool significantly reduces the time needed to investigate incidents across large surveillance networks.
icetana AI Analytics includes four core features designed to deliver actionable insights:
Yes, icetana AI includes a Facial Recognition feature that allows you to create and manage watchlists. The system can also detect and log faces in real time, providing actionable insights for enhanced security monitoring.
icetana AI’s License Plate Recognition (LPR) technology supports license plates that use Latin and Arabic scripts. This enables compatibility with a wide range of countries across different regions.
Yes, icetana AI includes its own License Plate Recognition (LPR) technology. It can detect and log license plates in real time, and match them against customisable watchlists to support proactive security measures.
icetana AI offers a range of products designed to enhance safety, streamline investigations, and improve operational awareness. Our product suite includes:
icetana AI is designed to detect key objects relevant to security environments, including:
The types of events detected by icetana AI depend on the product package you choose. However, common event types include:
Our team can help you choose the right package based on your specific surveillance needs.
icetana AI displays real time alerts on the Livewall, allowing operators to monitor events as they happen. For historical review, you can access previously detected events through the Highlights module.
icetana AI is designed to enhance safety and efficiency across a variety of sectors. Our core expertise includes:
Our platform is adaptable and scalable, making it ideal for any environment requiring proactive video surveillance.
icetana AI currently supports the following languages:
icetana AI provides comprehensive training tailored to different user roles. We offer online training, as well as remote or on-site sessions depending on the contract. Our dedicated Learning Management System (LMS) includes structured programs for end users, installers, and system administrators, each with role-specific training agendas.
It depends on the type of events you're looking for. Our Triage Agent lets you search for specific scenarios – like identifying if a knife appears in a scene – by using custom prompts. It works on top of icetana AI’s detected events, helping you quickly find what matters most to you.
icetana AI currently sends alerts to the Livewall interface, where operators can view and respond to events in real time. It does not send alerts directly to external devices.
With our core Safety and Security product, there’s no need to draw lines, our self-learning AI understands what’s normal in your environment. However, for features like Occupancy, Flow and Perimeter, you can manually draw lines to define key areas for alerts and analytics.
Events in front of the camera are shown to the Operators on the Livewall within 3 seconds of the event taking place.
icetana AI begins learning immediately and starts delivering alerts within 24 hours. Within a week, it becomes optimised to your specific environment. The system continuously learns what “normal” looks like, based on patterns, movement, and activity, and flags anything unusual. This adaptive learning runs year-round.
Yes, we've trained our core algorithms on over 700 million hours of video to recognise unusual activity in real time. It's adaptive, continuously learning what’s normal in each environment to reduce false alarms and improve detection over time.
Yes, the Safety & Security product is the foundation of the icetana AI platform. All other products, including Facial Recognition, License Plate Recognition (LPR), Triage Agent, Analytics, and Forensic Quick Find, are designed to work in conjunction with it and require it to function.
Currently, icetana AI does not process or analyse audio feeds. Its focus is entirely on visual anomaly detection and analytics.
icetana AI offers visual dashboards, event summaries, and downloadable reports through its Analytics and Highlights modules, helping stakeholders make data-driven decisions.
Return on investment can be measured through reduced operator workload, faster response times, fewer false alarms, and overall operational efficiency. For deeper insights, icetana offers analytics dashboards (available as an optional add-on) that highlight patterns such as high-traffic areas, useful for decisions around rent adjustments, staffing, and space planning. Our team can also help assess potential impact based on your specific setup.
Installation and setup time depends on the scale and complexity of your project. In most cases, it can take anywhere from 1 hour to 48 hours or more. Our team will work closely with you to ensure a smooth and efficient deployment.
Our Safety and Security product supports up to 400 cameras per server. If you're using additional products like Facial Recognition (FR) or License Plate Recognition (LPR), you'll need extra hardware to support those features.
Pricing varies based on the number of cameras and the product package selected. We’d be happy to discuss your specific needs and provide a tailored quote, feel free to book a meeting with our team.
Video processing with icetana AI can be performed either on-site or in the cloud, depending on your organisation’s preference. You have full control over the deployment method, and no external parties have access to your video data.
Currently, icetana AI does not support Single Sign-On (SSO) or Multi-Factor Authentication (MFA). We recommend implementing secure access practices within your organisation’s environment to manage user access.
Yes, icetana AI offers proof of concept (POC) licenses for organisations that are a good fit. This allows potential customers to trial the platform in a real-world environment before full deployment. Please note: POC licenses may be subject to a fee depending on the scope and requirements of the trial.
Yes, icetana AI is available with a perpetual licensing option, allowing you to make a one-time purchase for long-term use without recurring subscription fees.
icetana AI is licensed on a per-camera basis, allowing you to scale the solution according to the number of surveillance cameras in your system.
icetana AI is accessed via a secure web interface and is only supported on Google Chrome. To ensure optimal performance and compatibility, please use the latest version of Chrome when using the platform.
Yes, icetana AI supports multiple user sessions with configurable access levels, allowing teams to work simultaneously across different modules.
If your VMS isn’t listed as currently supported, please reach out to us. It’s possible that integration is already in development or on our roadmap. We’re always open to new opportunities and would value the chance to explore compatibility with your system.
icetana AI currently integrates with several leading VMS platforms, including Milestone, Genetec, and Nx Witness. If you use a different system, feel free to contact us, we’re continuously expanding our integrations and may already have support in development.
icetana AI logs camera disconnections and can notify operators through the interface when a feed is unavailable. Events will resume once the camera is back online.
Yes, icetana AI supports encrypted RTSP streams for added security. We recommend working with your IT team to enable encryption protocols.
Yes, icetana AI operates effectively at night, as long as your cameras can capture clear footage in low-light conditions. Our system analyses whatever your cameras see, whether it’s day or night. If the scene is visible to the camera, icetana can detect and analyse it.
Yes, users with the admin role can do this.
No, icetana AI can operate without a Video Management System (VMS). While a VMS is not required, icetana is fully compatible with a wide range of VMS platforms. This means you can use icetana as a standalone AI surveillance solution or integrate it seamlessly with your existing VMS to enhance your security operations.
No, we're largely camera agnostic and compatible with most existing systems.
Yes, icetana AI supports 360-degree cameras by processing each lens feed individually. Most 360-degree cameras use four separate lenses, and each lens is treated as an individual video stream. As a result, you would need to license each lens separately, meaning four icetana AI licenses for a typical 360-degree camera.
While icetana AI can technically work with PTZ cameras, we don’t recommend it, especially for our self-learning technology. When a PTZ camera moves, the system has to relearn what’s “normal” in that new view, which can reduce effectiveness.
icetana AI is optimised for standard visible-light video and is not officially designed to support thermal or infrared cameras. However, it may still function with these camera types if the footage is clear and meets certain visual criteria. Please note that performance and detection accuracy cannot be guaranteed when using thermal or infrared video.
The required frame rate depends on the specific icetana AI product being used.
For optimal results, we recommend the following video stream resolutions:
To ensure a successful installation and optimal performance, the following prerequisites are typically required:
Our team will provide a detailed installation checklist and work closely with your IT or security team to ensure all requirements are met.
icetana AI is compatible with Ubuntu Server 22.04 LTS and Red Hat Enterprise Linux 8, ensuring reliable performance on industry-standard Linux environments.
Yes, additional sites can be added at any time after deployment. icetana AI is built to scale with your needs. As your surveillance footprint grows, you can easily integrate new sites into your existing setup. Depending on the volume of video feeds and system load, expanding may require additional hardware to maintain performance and efficiency. Our team can provide recommendations to support smooth scaling.
No, icetana AI does not place a limit on the number of sites you can add. Our system is designed for large-scale environments and is ideal for organisations managing multiple locations. Video feeds from all sites can be streamed to a centralised location for monitoring and analysis. The number of supported sites ultimately depends on the capacity of your chosen hardware infrastructure, and we provide guidance to help ensure optimal performance at scale.
Yes, icetana AI supports domain-joined server configurations. This allows integration with your existing IT infrastructure and policies, including centralised management and authentication controls.
At this time, icetana AI does not offer built-in support for high availability or redundancy. However, we can provide guidance on how to architect your deployment for improved resilience using your existing infrastructure and tools.
We recommend NVIDIA RTX 4000 Ada for optimal compute power per PCIe slot.
icetana AI runs on standard, commercially available server hardware that meets our recommended specifications. We don’t sell customised hardware, allowing you to purchase from your preferred certified providers or resellers. To support this, we provide detailed hardware guidelines to ensure optimal performance, including processing, storage, and GPU requirements where needed.
If you're deploying on-premise, you’ll need to provide suitable servers equipped with GPUs to run icetana AI efficiently.
No, icetana AI does not provide its own cloud hosting services. However, you can deploy icetana in your preferred cloud environment, such as AWS. This allows you to retain full control over your infrastructure while leveraging the scalability and flexibility of the cloud.
icetana AI’s Facial Recognition feature is compatible with body-worn cameras. While the core anomaly detection features are designed for fixed surveillance systems, Facial Recognition can be used with video footage captured from body-worn devices, provided the footage meets the necessary quality and angle requirements.
Yes, icetana AI supports cloud-based deployments. Our platform is flexible and can be deployed in the cloud to suit organisations that prefer scalable, remote infrastructure. Whether you're operating across multiple locations or aiming for easier maintenance and updates, our cloud deployment option offers a secure and efficient solution.
Yes, icetana AI fully supports on-premise and dark site deployments. Many of our customers choose this option to maintain complete control over their security infrastructure. By deploying icetana AI on-premise, without any external network access, they ensure that sensitive video data remains entirely within their secure environment. This is particularly important for high-security sites where privacy and access restrictions are critical.
.webp){kind=logo}