AI Toolkit

All types of AI Appliances and AI Boards are created through the integration of Gorilla’s AI Toolkit — comprised of video analytics and OT security AI models.

Gorilla’s AI Toolkit empowers clients to leverage and combine Gorilla’s AI platforms with hardware or IoT devices of different form factors – this shortens time to market for end product manufacturers and solution developers.

Video Analytics AI Models

Intelligent Video Analytics (IVAs) are sets of AI models that use AI algorithms and metadata to scan video in order to detect patterns or recognize specific objects. Once processed, the video data can then be queried and searched for various/specific results.

Gorilla is a leader in the market by pairing our video analytic specialization of over 40 IVAs with leading industrial PCs, net appliances, vehicle PCs, and servers from partners like Dell-EMC, Lanner Electronics, and Neousys. Considering the digital transformations COVID-19 has spurred many organizations to undertake, the delivery and quick deployment of cutting edge AI Appliances that focus on physical, data, and network security has proven invaluable.

Video is processed to generate analytics based on the following categories:

View Video Analytics AI Models by Vertical

OT Security AI Models

As industrial systems become more connected to build powerful, networked businesses, they also become more exposed to cyber threats. Connecting an OT network to an IT one instantly exposes the OT network and all connected OT devices to the entire threat landscape. Ever-evolving malicious attacks and other risk factors stemming from the fact that some endpoints are operating on outdated software lead them to be more susceptible to hackers. In addition, the rise of remote access to OT networks by third-party vendors further expands the attack surface and creates new vulnerabilities. The high cost of industrial equipment and the devastation to businesses and communities that a cyber-attack could generate are key factors for organizations looking to protect their industrial networks.

Security convergence is the melding of different IT and OT technologies into an interconnected system, spurred on by digital transformation. Having an effective security convergence strategy in place and ensuring endpoint protection will protect processes, people and profits while significantly reducing security vulnerabilities and incidents. Security convergence gives organizations a single view of industrial systems together with process management solutions that ensure accurate information is delivered to people, machines, switches, sensors and devices.

To protect customer’s data and devices, we offer a wide-range of data, network, and endpoint security solutions to ward off attacks, along with AI-based analytics tools and comprehensive data monitoring and alert systems.

Network Anomaly Detection

Network anomaly is a state or condition in which network operation deviates from normal. Many network security devices use rules or signatures to compare network packets for the detection of known anomalies. We use AI technology to learn various abnormal behaviors on the network to generate AI models for network anomaly detection. These AI models can be used to detect various abnormal network behaviors, such as denial of services, network topology anomaly, abnormal network traffic, abnormal communication protocol, network intrusion, etc.

Malware and Suspicious Behavior in Endpoint Environment

Traditional antivirus software mostly uses signatures or rules to detect malicious programs in the endpoint environment. The advantage of this approach is that it can quickly and effectively detect malware, but the disadvantage is that the detection rate for unknown malware is very low.

We collect a large number of malicious programs and footprints generated by various malicious behaviors in the endpoint environment, and use Deep Learning technology to train AI models that detect malicious programs and behaviors. The main advantage of using these AI models is that they still have a good detection rate when facing unknown malicious programs or behaviors; in addition, there is no need to frequently update virus patterns or detection rules.

View OT Security AI Models by Vertical