Red Hat’s OpenShift delivers AI at the edge
The two biggest fronts that federal IT modernization are currently pushing into are artificial intelligence/machine learning, and edge computing. Federal agencies need to process their data faster, and they need to do it closer to where the data is collected and used. Delivering this level of sophistication at the Edge has been challenging, along with keeping the approach and technologies consistent with what’s used in data centers and the Cloud.
“With Openshift — built on Kubernetes — we believe we have the right platform for data science workloads wherever they’re needed. We’ve recently extended that story to the Edge, which is really exciting news for our customers,” said Eamon McCormick, senior manager specialist for emerging technologies.
Red Hat recently delivered updates to OpenShift Container Platform that enables deployment of a small, 3-Node footprint. This architecture has been validated with hardware partners, including HPE and NVIDIA, for delivery via ruggedized, edge computing platforms. McCormick said it enables datacenter-level capabilities to be delivered into hospitals, ships at sea, aircraft, vehicles and other remote facilities.
“There are intelligent applications that need to be run by the government in situations where network latency just can’t exist,” McCormick said. “Applications that are running on that hardware are essential to groups that operate in the field. Processing data and running models and intelligent applications at the edge delivers faster, more reliable service to critical missions.”
This approach can be applied to many government focus-areas where decisions have to be made in real-time with a high degree of accuracy. Where this speed is required, or where connectivity simply isn’t available, communicating with Cloud or data center services simply won’t work.
Hospitals can use this solution to intelligently assign patients, staff and rooms. They can use the history of a patient, any past trauma or mental health issues, to predict susceptibility to future problems and act preemptively to improve patients’ lives. Federal law enforcement agents can process data in the field in real-time to prevent attacks, or simply protect themselves in dangerous situations. Social service and financial services agencies can better detect fraudulent enrollments, claims, and other activities that cost the country billions of dollars annually.
As always with Red Hat, this offering is built on open source software and open design principles. That means customers can take advantage of the innovation happening in open source communities, but also plug in commercial technologies as part of a comprehensive approach. There’s not a one-size-fits-all solution for data science at the Edge, and this open approach allows agencies to build the right solution to support their specific mission.
“We have over 140 partners who have now certified solutions to run in a fully automated manner on OpenShift. They are building Linux images on our RHEL Universal Base Image, along with Kubernetes Operators to run on OpenShift,” McCormick said. “The Operators built for their software components automate the deployment and management of those technologies running on Openshift. A lot of our AI/ML partners have either completed the certification program or are in process at this moment. Their participation really expands the options our customers have to address real mission challenges.”
That’s important because the edge is the least developed stage of hybrid cloud, which is the IT modernization model that most federal agencies are moving toward. Most agencies have data centers, a footprint in the public cloud, and remote offices/vehicles/stations. Many are leveraging Internet of Things sensors and devices for data gathering in the field and supporting real-time decision making.
“The Edge is just another footprint of the hybrid cloud model. Data science at the edge cannot happen successfully in a silo. Extending OpenShift to the edge is the natural evolution of the platform itself and it enables consistent DevSecOps practices from the data center, to the Cloud, and to the Edge,” McCormick said. “Along with supporting all development and delivery aspects, the platform is ideal for operating and scaling the workloads after they’ve been deployed.”
“Red Hat is providing the plumbing and electricity for the hybrid cloud,” McCormick said. “OpenShift, Container Storage, and Red Hat Application Services can now be deployed anywhere. That gives our customers portability, consistency, and flexibility wherever they operate.”