Plug Criterion AI right in
Most IT environments of pharmaceutical companies are rather complex and having multiple systems from different vendors communicate with one another is often no easy task. We know that and, therefore, we have invested in making Criterion AI incredibly simple to integrate. That means that implementing Criterion AI into existing IT environments can be done in just a few days—not weeks or months. 👏
We support many on-premise systems such as relational databases from vendors like IBM, Microsoft, and Oracle as well as industrial solutions from ABB, B&R, Beckhoff, Omron, Rockwell, Siemens and many more. All of our interfaces are based on industry standards, which means that integrating Criterion AI with other systems is not a matter of developing custom components but rather configuring a few settings through our intuitive user interface.
In addition, we provide interfaces to the leading public cloud providers in order to easily import and export data to and from Criterion AI. We expose data stored in Criterion AI via an Amazon S3-compatible interface as well as an FTPS interface using explicit encryption via TLS, which both offer enterprise-grade performance and security. 🔐
RESTful APIs and OPC UA TSN
AI models trained in Criterion AI can both deployed in the cloud and on local hardware without connection to the Internet. Online deployments hosted by Google Cloud Platform’s AI Platform for Tensorflow and are exposed via a secure RESTful application programming interface (API). REST is an industry standard for interacting with online resources over HTTP and is an acronym for Representational State Transfer. Our RESTful interface makes it a breeze for your application developers or external collaborators to make use of your AI models trained in Criterion AI. 💨
For offline deployments, we support the OPC Unified Architecture protocol over Time-Sensitive Networking (also known as OPC UA TSN). It is a new and powerful protocol for communicating between servers and clients on a network over OPC UA in real time. While OPC UA TSN is still relatively new, it enjoys support from all of the major automation equipment providers in the world. To find out more about the support for OPC UA TSN, check out the OPC Foundation’s press release from November 2018.
If you would like to learn more about OPC UA TSN or if you would like to discuss the integration of models trained in Criterion AI over other protocols such as EtherCAT, EtherNet/IP, PROFINET, POWERLINK or others, please reach out to us. We would love to talk to you about your integration needs. 🙂