A few days ago, the Wall Street Journal ran a piece about GE Digital’s investment in IOT hardware devices in order to improve the edge computing capabilities of its Predix platform. From GE’s perspective, integrated hardware devices can help to address the “last mile problem” about integrating PRedix with intelligent industrial equipment.
The race for dominance in the IOT platform market is becoming a contest between two main groups of technologies. On one side we have platform as a service(PaaS) providers such as AWS, Azure and Bluemix that have added native IOT services to its platforms. The other category of technologies includes industrial powerhouses such as PTC or GE Digital that have released outstanding IOT offerings. While the IOT PaaS providers are focused on providing robust horizontal IOT platform capabilities, the industrial vendors are banking on their domain expertise to build capabilities and solutions that improve the viability of their platforms in industrial environments. To both groups of technologies, edge computing and the integration with industrial devices is one of the biggest challenges in order to streamline the adoption of their IOT platforms.
Without exception, all the top IOT platforms in the market have developed some form of edge computing mostly on the form of device discovery and management gateways. While those capabilities are certainly relevant, they are hardly sufficient to address the challenges of integrating sensors on an IOT industrial environment with a specific IOT platform. In those scenarios, the integration of legacy industrial equipment with modern IOT platforms requires a monumental amount of work. That process is also hard for IOT device manufacturers that are constantly looking to make their devices work with the new generation of IOT platforms.
In order to address some of the aforementioned edge computing challenges, GE Predix has been building hardware devices that integrate with legacy industrial equipment and broker the communicating with the instances of the Predix platforms running on that industrial environment., GE’s solution also includes other relevant edge computing IOT capabilities such as storage and messaging.
Solving the “last mile” integration on IOT topologies is an absolute requirement to streamline the adoption of IOT platforms in industrial environments, However IOT edge computing solutions are far from trivial and should include some really sophisticated capabilities in order to be effective. Let’s explore a few of those capabilities that are required in order to streamline edge computing IOT solutions.
— Device Adapters: IOT Edge computing solutions should provider hardware-software adapters to collect data from sensors and devices in industrial environments. Adapters should support IOT protocols such as MQTT, XMPP, etc.
— Data Caching: IOT Edge Computing solutions should enable the caching and storage of data generated by IOT devices.
— In-device Computing: IOT Edge Computing solutions should provide local, in-device computation capabilities. Technologies such as AWS Greengrass are a good model for that type of solution.
— Bidirectional Communication: IOT Edge Computing solutions should support bidirectional communication between device adapters and the main IOT platform.
— Device-to-Device Communication: IOT Edge Computing should enable autonomous communication between devices on an IOT network as well as with the centralized IOT platform.