Building Smart Cities with AWS Kinesis
The Cloud accelerates exciting new innovations, making it easier for entrepreneurs to develop, build and launch new products much faster in hot sectors like the IoT (Internet of Things) and how it can be applied to scenarios like Smart Cities, as Amazon describe in their solution set and blog series.
For Scotland the field offers a potent goldmine of opportunities, as they set out to pioneer their own smart cities, a process that will expose multiple niche segments for entrepreneurs to fulfil, as demonstrated by examples like Irish startup Wia. Scottish entrepreneurs can do the same!
From batch to real-time
It’s not a point solution specifically for Smart Cities rather a tool set that enables developers to build any type of application that processes large volumes of data streams, like sensor networks deployed around a city for monitoring all of its operations, from traffic tolls through garbage levels. As AWS describe this same capability could also be used for click streams from web sites.
The key technological leap is from ‘batch’ processing, how IT architecture has traditionally dealt with handling very large volumes of data where it is stored up over time and then loaded into analytics tools in one job, to a real-time mode, where it is handled as it happens, as a stream of data.
The Zillow case study describes how the real estate business utilizes machine-learning calculation performance and scalability to deliver near-real-time home-valuation data to their customers. Netflix uses it for real-time monitoring of their vast infrastructure.
IoT Pragma Architecture – Small things and the Cloud
In their best practice white paper AWS provides a detailed overview of how to implement Kinesis, and in Core Tenets of IoT they define a ‘Pragma Architecture’ for building IoT solutions. This defines the component parts of building an overall IoT solution such as:
- IoT Device Gateway: The interface between the devices and the processing applications, using protocols such as MQTT, WebSockets or HTTP.
- Device Registry: A directory system for registering each of the devices and issuing them x.509 certificates, to enable secure communications via TLS.
- Event Driven Apps: Middleware for enabling app integration, utilizing the SQS – Simple Queue Service and SNS – Simple Notification Service.
AWS Solution Architect Brett Francis gave this presentation that explains the detail of this blog, offering a primer for ‘Small Things and the Cloud’ and exploring the challenges of these telemetry applications.
IoT scenarios also provide an ideal context for ‘Serverless’ applications, an approach that can be implemented through the AWS Lamda service.
In essence it represents the pinnacle of Cloud computing because as the name suggests the key benefit is to abstract developers entirely from the underlying machinery of servers, freeing them to focus purely on adding value via the apps that they create.
It also works on an entirely utility model, Lamda executing code when it is needed and billing only for that consumption, ideal for the real-time, highly variable requirements of IoT scenarios, as AWS describe in this guide.