Being digital is what enterprises of all shapes and sizes aspire to today. It allows them to be responsive, contextualized and instantaneous. In short, it empowers them to act organically and respond intelligently to market changes. Humans can respond quickly due to the advanced capabilities of our brain. For instance, the cerebrum receives and processes sensory information, controls language and communication, and is responsible for learning and memory. Based on these inputs and analyses, we respond appropriately to external stimuli. Organizations should have a similar unit focused on functions across their enterprise. Let’s call it the Enterprise Cerebrum.
The Enterprise Cerebrum could receive and process data/inputs, look at historical data and context, and enable leaders to make intelligent decisions in real-time. It could also use the data to continuously learn and improve organizational efficiency. For example, consider fleet management as a tangible use case.
As organizations cope with rising demands for faster delivery at competitive prices, it becomes imperative to maintain and manage a fleet with optimal performance. This requires the ability to perform dynamic rerouting based on traffic or road conditions – while maintaining a real-time view of the truck’s health – to find the most economical route.
Trucks have telematics systems for geofencing, as well as sensors for measuring temperature, speed, fuel, capacity utilization and driving behavior. These are complemented by the onboard computer, which reports the truck’s health. This is transmitted through the mobile network and received as events through an Event Mesh.
Data analytics and machine-learning models can operate on the events to extract information and reroute the truck or issue commands to the driver in real time. They can also use the data about capacity and load to optimize the route recommendations, while business stakeholders can view and propose a different vehicle or other route changes. In the case of lost or stolen trucks, the enterprise can identify driving patterns from the data and route information, determining in real time whether it’s prudent to remotely disable the vehicle.
All of these functions of an Enterprise Cerebrum help the organization make fleet-management decisions and maintain smooth business operations. Although it sounds complex, systems such as this can work effectively if designed to focus on receiving, processing and drawing inferences from organizational inputs we’ll call “events.”
Enterprises typically receive inputs from channels such as mobile, web, cloud, devices and sensors, human actions, and changes to applications and processes. These inputs, or events, include business notifications like orders being placed or shipments arriving, sensor notifications such as speed or temperature changes, infrastructure events like hardware status updates, or application notifications such as user logins and module changes. These inputs are typically published as messages to the Enterprise Cerebrum.
An Event Mesh is an architecture pattern that receives these inputs and enables the enterprise to process them using a messaging layer and event-stream layer. It also includes the infrastructure to implement a messaging mechanism that can be deployed on the cloud or on premise and provisioned on-demand. Using the Event Mesh, organizations can gain a centralized view of the events flowing across the application infrastructure, with receiving applications notified of the events’ arrival.
The Enterprise Cerebrum enables businesses to process inputs via real-time analysis and historical data analysis. Some Event Mesh products allow functions to be associated and executed in real-time upon an event’s arrival. The functions could be deployed as code on premise, on the cloud and as containers; or they could be deployed on the cloud on Serverless platforms that charge only for the time the function ran and/or for the resources it consumed.
“Cloud Events” is a new event-format specification being drafted to enable consistent event processing across cloud platforms, IoT devices and other middleware. With a common event format, events generated from an application on one cloud provider can be consumed in a function or application built on another cloud platform, enabling Function-as-a-Service providers such as AWS Lambda and Azure Functions to interoperate. This common format also allows consuming, correlating and tracing events across IoT devices and other middleware, enabling functions with business rules to take action automatically based on historical data and context.
In addition to real-time analysis, the Enterprise Cerebrum can perform historical data analysis by reviewing the event history, context and then correlating them to take the appropriate action. Analytics engines such as Apache Spark use event data to identify patterns and make inferences, which helps prioritize events to handle the volume of events correctly, generally through AIOps. The analytics engines work in tandem with data warehouse platforms to process large data sets, with the platforms providing a standard mechanism such as SQL to access the data, define and automate tasks, employ statistical models on the data and visualize it.
Drawing Inferences and Decision Making
Enterprises can leverage Machine Learning models to make sense of the data gleaned across channels, identify relationships and analyze sentiments, ultimately aiding in their decision making. OData, an open protocol, can be used to make RESTful API calls and add context and meaning to the data through semantics, making data queries and modifications more understandable.
The foundation of the Enterprise Cerebrum is making sense of events and data from multiple sources, correlating and understanding the context, and planning and acting accordingly. This conceptual architecture mimics the human brain to help enterprises respond to events in real time and take strategic actions based on them. This solution of the future could function both independently or as a means to help stakeholders with planning and making informed decisions. In essence, the Enterprise Cerebrum would be the entity that continuously learns and adapts, helping to enable a truly smart enterprise.