Continuing in unstoppable mode when it comes to participating aggressively in the world of connecting things, Amazon has rolled out their AWS Greengrass solution enabling developers to run Lambda functions at the edge, on IoT devices, while also making it easier, according to Amazon, to communicate with each other and synch with the cloud.
Keeping computing local and stateful, Greengrass was built on the combination of AWS IoT and AWS Lambda, and promises efficient local computing, cloud native connectivity, local messaging between endpoints, and access to a range of AWS services.
In the spirit of delivering a developer friendly experience, and low-code integration, three Software Developer Kits (SDKs) are now generally available:
Core: includes APIs that invoke Lambda functions, connect with the cloud, publish messages to local device networks and keep data in state
AWS IoT Device: a connector supporting TLS, WebSockets, and MQTT, with devices to AWS IoT as well as C++, Java, Node.js, and Python languages.
AWS Python: used to write Lambda functions that interact with AWS services.
Greengrass Groups
Local networks comprise any number of devices, or IoT endpoints, that have one or more SDKs installed, which are then provisioned on a local network with local compute and connectivity. With minimal requirements met, devices can support Over-the-Air (OTA) updates for anything from a sensor to a smart appliance, wearable devices and more as long as they include micro-controllers with the AWS IoT device SDK implemented.
AWS Lambda is the company’s “serverless” processing service. Serverless computing (also known as function as a service) is a cloud computing code execution model in which the cloud provider fully manages starting and stopping of a function's container platform as a service (PaaS) to serve requests, billed by an abstract measure of the resources required to satisfy the request, rather than per virtual machine, per hour.
Serverless code can either be triggered by specific events (including user registration with Amazon Cognito), or be configured to run behind an API management platform in order to expose it as a REST API endpoint, which is the case with Greengrass.
Two Integrations, More to Come?
Lenovo and Bsquare both announced integrations with Greengrass timed with AWS’s official announcement.
Lenovo will pair its network of IoT providers with devices to analyze data received from IoT sensors, taking advantage of Lenovo’s offering combined with Greengrass services. Lenovo is the “preferred endpoint partner” for Greengrass in North American and the UK, with the first device tested and certified their ThinkCentre M Series710 Tiny Desktop with Ubuntu, and will analyze and store IoT data at the Edge.
“Lenovo has a unique position in our industry because we can deliver the devices, smart connectivity and infrastructure required to create a great user experience and satisfy user needs,” said Jon Pershke, vice president, Business Group Strategy, Lenovo, in Lenovo’s announcement. They are as much of a sales channel partner as a technology partner, helping commercialize Greengrass, AWS’s preferred go-to-market for most services.
Lenovo claims AWS IoT customers can reduce their total cost of ownership with “reliable, configurable and low cost PC end point devices for real-time data processing, and have their sights set on industry verticals including healthcare and retail. Working with Greengrass is related to their strategy to play in the growth market of computing at the edge (while also being able to send data to a centralized cloud for deeper analytics and decision making for geographically distributed deployments).
Bsquare also participated in the Amazon Web Services (AWS) invitation only Limited Preview program, and announced their DataV platform is now fully interoperable with Greengrass. Utilizing advanced analytics models built from device data, DataV claims their software platform can auto-generate AWS Lambda functions which are then executed on AWS Greengrass Core enabling edge analytics without the need to program AWS Lambda functions directly.
"By using the newly announced AWS Greengrass service, DataV further enables a distributed compute configuration where raw data is processed at the point of ingress and only relevant information is passed back to the AWS cloud via the Device Shadow mechanism," said Dave Wagstaff, Bsquare Chief Technology Officer. "This distributed compute model allows companies to realize benefits that include near real time event processing on edge devices, reduced data transport costs and quick implementation of edge based DataV applications."
Intel Inside, Qualcomm
AWS announced Greengrass at its re:Invent conference in 2016, and seem to have a lead on edge processing compared to their competition (for example, Microsoft Azure).
At that time, Intel announced at the end of last year a new IoT Dev Kit and Greengrass compatible gateways enabling hardware manufacturers and enterprises adopt the new services.
Qualcomm reiterated their support for Greengrass in select system-on-chips (SoCs). Manufacturers expected to commercialize solutions running AWS Greengrass on Qualcomm Technologies’ chips include Advantech and Thundersoft, with commercial products expected to be available later this year.
Using multi-protocol communications, SoC powered gateways can aggregate data and support a wide variety of field sensors and devices, including legacy equipment with no native IP connectivity. Local data processing and control signaling in gateways helps reduce overall system latency, while local data storage allows for enhanced security and privacy using the security features available in Qualcomm Technologies’ chips.
The Greengrass cloud service is generally available out of AWS’ Northern Virginia and Oregon data centers. The service will be made available from the cloud provider’s Frankfurt and Sydney regions in the coming weeks.
