By definition, realtime data is information that is delivered immediately. Older legacy systems that claim to support realtime in fact do not. But even modern cloud-based systems, despite their support for easy connectivity via APIs and mobile scenarios, were not designed for the contemporary take on realtime.
A common root cause for a system to fail delivering true realtime is because of its limited or inadequate, underlying architecture. Contemporary integration uses webhooks to receive notifications when an event occurs. While this is generally accepted as ‘close enough’, you’re actually just retrofitting an existing, inadequate architecture and in the process of patching up one issue, creating other problems.
What is encouraging is that some components of newer systems are being designed to meet the needs of IoT (Internet of Things) – including intense data workflows, support for smaller software footprint on devices with limited compute power – and there are glimmers of true realtime scenarios among them. However, you also need integration technologies to support realtime properly to build real solutions that are useful in the real world.
IoT is a good place to examine realtime deployments, even if you don’t have an IoT project on your ‘To Do’ list because a successful IoT experience relies on enabling data to flow in realtime. When true realtime data can flow back and forth, actual business goals can be met, such as providing better customer service, improving the effectiveness of loyalty programs or increasing time to market with a new offering.
In this article I’ll reveal a pattern for delivering realtime data using a surprising ingredient – cloud integration.
A Recipe For Success: An Integration Pattern That Supports Realtime
The most common realtime integration recipe almost always includes the following components: data stream processing, rule matching and integrating with other systems to provide some kind of notification to an end user, woven together with an integration Platform-as-a-Service (iPaaS) and, at times, a Mobile Backend-as-a-Service (MBaaS).
The benefit of this pattern is that it allows you to plug and play into different systems and use cases. The ingredients to the solution remain the same no matter the context; be it mobile, web, or IoT, consumer or industrial, prototype or large-scale deployment.
Although business goals differ depending on the organization, a properly designed ‘connected architecture’ (and the workflows that support it) enables the use of a consistent pattern that can be recycled to solve very different business challenges.
To illustrate this point further, here are some variable use cases where realtime data delivery is expressed as different flavors to a reusable pattern.
Use case #1: Loyalty programs and sales in connected spaces
Sales and loyalty programs are only useful if they play out in a timely manner. It’s rare that a customer will remember to bring a coupon they see in a newspaper to a store, but if he or she receives a notification for a coupon on their smartphone as they walk toward a store, they might actually walk inside and use it. Why? Because it’s timely, convenient and highly contextual.
Under the surface, the technology that helps guide customer behavior is, of course, realtime data. In the context of a connected space that incorporates a loyalty program, the integration workflow might look like the following:
Data is collected in the MBaaS from an IoT device like a beacon, sensor or mobile app. This essentially functions as a database, where it can be easily accessed by a number of other services through proper integration. This means if a space is “connected” it can access the loyalty system, the location-based couponing service and the self-checkout system all from one location using integrated data.
You can copy and paste this integration pattern into a variety of scenarios, like connected hospitality, connected retail or a connected event space.
Use case #2: Preemptive customer service (for physical products as part of the Industrial Internet of Things)
Great customer service relies on predicting customer needs. In this world, nothing beats preventing or solving an issue before a customer even experiences it. Once again, the technical recipe to help support this means access to realtime data streams from multiple services and locations. Here’s an example of an IIoT elevator use case to help explain this pattern:
IIoT brings real value to large scale technology deployments – such as a bank of elevators in a commercial building – by reducing the cost of maintenance, optimizing spare part stocking and increasing overall service efficiency. An elevator with intelligent monitoring systems can alert a company (or a person) of potential issues before something actually breaks down. From there, the “smart” elevator can show the service technician the exact part that needs to be replaced and automatically deliver the service manual to the person’s mobile device--all before someone ever gets stuck in a broken elevator.
In the context of a physical product / industrial IoT scenario, an integration workflow might look like this:
Sensors attached to an IIoT device can funnel data into your database or MBaaS, and then channel it into an AI system that can compare the status of the device to historical data and predict if there will be an issue. If a problem arises, based on business logic or rules, the ticket can be routed to the right service agent so they can proactively work to solve the problem before the customer is negatively affected.
Again, this integration pattern could be used in a variety of IIoT contexts, ranging from vending machines, to jet engines, to supply chain management, to connected transportation.
Use case #3: Engineering feedback loop for connected transportation in a post-production supply chain environment
Continuous improvement is one of the best ways to get ahead in IoT development. And a key component to help set a company up for success is realtime data integration. In the context of connected transportation, an integration workflow might look like this:
Modern vehicles provide a constant stream of realtime data, which can be used to streamline the post-production maintenance and diagnostics of a connected vehicle, such as a Tesla.
For example, if a vehicle’s brakes are overheating from normal usage, a notification could be sent automatically to the car’s owner to take it to the nearest service center via directions through an integration with Google Places and Waze. The car would then notify the service center that it’s on its way with the estimated arrival time. From there, the parts center would be alerted to send a delivery to arrive at the same time as the car itself.
The value of integration and realtime data in a post-production supply chain is that it reduces the lifetime cost of supporting human lag time. Instead of a vehicle sitting in a service center for days while the car’s problem is being diagnosed and parts are delivered and installed, with realtime data integration, the vehicle can be back on the road later that same day, instead of a week later.
By providing car technicians with pertinent information in realtime, they’re able to do their jobs better and more quickly, ultimately positively impacting the customer’s experience. Implementation of realtime technology not only improves customer satisfaction, but also reduces cost and trouble for the dealership, manufacturer and owner.
The Future Of Realtime
Connecting systems to manage realtime data flow used to be incredibly finicky and required specialized knowledge and tools. More recently, industry standards and protocols have allowed such integrations to become more standardized.
Similarly, a new generation of cloud integration platforms are making it faster, easier and cheaper to orchestrate the necessary connections: A growing number of innovative iPaaS (integration platform-as-a-service offerings) provide the flexible glue to connect any service with quasi-realtime support today and replace it with a true realtime system as it becomes available. This convenient plug-and-play framework allows for a continuous upgrade cycle, where realtime capabilities can be added or upgraded on-demand, without breaking the overall architecture or solution.
With an iPaaS, you can integrate terabytes of data from disparate cloud systems, build triggers that kick off data workflows and maintain connections to true realtime systems and devices, including sensors and beacons. As the costs associated with IoT continue to decrease, I expect realtime data flows to increase exponentially everywhere around us. In addition, iPaaS also makes it possible for a new generation of non-technical users to create realtime data flows.
Imagine empowering the facilities team in an office building to create automated workflows themselves. Or allowing security to connect realtime building access data with live video feeds and a facial recognition service. Such solutions would have taken months to build not that long ago and would not have been realtime. We’re at a point in time where even a semi-technical user can create solutions for themselves in a fraction of the time and without specialized knowledge.
The future is exciting, thanks to integration coupled with new technology. Soon most scenarios we imagined as futuristic will feel normal.
In another example, consider the City of San Francisco: five years ago bus schedules were proudly displayed over many bus stops, but they showed when the bus should be there versus when it would be there. Once an API was released, apps like NextBus were able to process realtime data to show exactly where the buses were and how long it would take for it to actually arrive. Today, this feels expected.
Now, think of a world where a city’s transportation system, its traffic monitoring system, its traffic control system, weather information and private solutions such as Lyft and Google Maps can all share and compute data instantly. The system can analyze and reroute traffic based on current and future conditions, which represents a major leap forward compared to what was possible by relying on machine learning based solely on historic traffic patterns and near-realtime data.
The future is bright. It’s connected. And it’s realtime.