All News | Boards | Chips | Devices | Software | Archive | About | Contact | Subscribe
Follow LinuxGizmos:
Twitter Facebook Pinterest RSS feed
*   get email updates   *

Amazon's AWS IoT platform taps three Linux SBCs

Oct 9, 2015 — by Eric Brown 10,096 views

Amazon’s new “AWS IoT” cloud IoT platform offers Starter Kits built around Linux-ready SBCs like the BeagleBone Green, DragonBoard 410c, and Intel Edison.

Amazon made its first big Internet of Things play by launching an IoT managed cloud platform for aggregating and processing IoT endpoint data, built around its Amazon Web Services (AWS) platform. Available now in beta form, AWS IoT, is being made available in the form of a series of AWS IoT Starter Kits, which bundle popular hacker boards with the AWS IoT Device SDK, and in some cases other hardware such as Grove sensors. Three of the 10 kits runs Linux, including kits for the DragonBoard 410c, BeagleBone Green, and Intel Edison (see farther below).

BeagleBone Green (left) and Intel Edison with Arduino AWS IoT kits, both with Grove sensors
(click images to enlarge)

AWS IoT enables users to “store, process, analyze, and act on the volumes of data generated by connected devices on a global scale,” says Amazon. Devices connect to AWS IoT Device Gateway, using both HTTP and Message Queue Telemetry Transport (MQTT), secured with mutual authentication and encryption. The gateway integrates with AWS Identity and Access Management (IAM) to let customers set granular permissions for individual devices, or fleets of devices, and manage them throughout their lifecycles.

With the AWS IoT Device Gateway, manufacturers can set rules for how AWS IoT handles and responds to data triggers from their devices. For example, they can script actions such as sending an alert when a motion detector is triggered, or when a pressure sensor reports an unusually high reading, says the company. You can create rules that apply to data from a single device, a group of devices such as a sensor array, or a mix of devices and data sources, such as data stored in Amazon DynamoDB. In addition to DynamoDB, AWS IoT provides integration with Lambda, Amazon Kinesis, Amazon S3, and Amazon Machine Learning.

AWS IoT conceptual diagram
(click image to enlarge)

In the case that devices lose access to power or the Internet, AWS IoT creates a virtual “shadow” profile of each connected device using REST APIs. The shadow includes all of the information about the device’s state, so applications can continue to check device status or take actions. These are automatically sent to the device once it reconnects.


An AWS IoT Management Console offers a web-based interface for managing these processes, including tasks like creating a “thing,” associating certificates, publishing messages, and defining actions, says Amazon. The console interacts with both the AWS IoT Device SDK, and the cloud-oriented AWS SDK, which provides APIs for applicable AWS services.

An AWS Partner Network (APN) program includes companies like Micrium and Canonical, whose Ubuntu Linux distribution, as well as Micrium’s μC/OS RTOS, were most prominently mentioned as OSes that support AWS IoT. Others include Ayla Networks, Cirrus Link, Thingworx, and Xively

According to a TechCrunch story that covered the announcement at Amazon’s re:Invent developer conference in Las Vegas, the technology is based on the company’s acquisition of IoT development firm 2lemetry earlier this year. AWS IoT will be sold as a pay-as-you-go service, with the first 250,000 512-byte message blocks per month free, and $5 per million message blocks after that, says the story.

The story notes that the platform will compete with other cloud-oriented IoT device gateway services such as Microsoft’s Azure IoT Suite. Other cloud-oriented gateway solutions include Intel Gateway Solutions for IoT.

AWS IoT Starter Kits

Amazon announced 10 AWS IoT Starter Kits, three of which run Linux and in the case of the DragonBoard 410c, Android. The rest run Arduino or other microcontroller-based hacker boards.

The three Linux-ready kits all use the C and Node.js components in their integrated AWS IoT Device SDKs. Two of the kits — BeagleBone Green and Intel Edison — also include Grove sensors and actuators. Seedstudio, which manufactures the BeagleBone Green, is quickly turning its Grove product line into de facto standard for IoT hacker kit sensors and add-ons.

The three Linux-ready AWS IoT Starter Kits include:

    BeagleBone Green
  • BeagleBone Green and Grove AWS IoT Starter Kit — Seeedstudio’s $99 BeagleBone Green and Grove kit is based on the now-shipping, $39 BeagleBone Green. The BB Green is not exactly the successor of the BeagleBone Black, but it has the full support of the community.

    The $39 SBC subtracts the BeagleBone Black’s micro-HDMI port, but adds a micro-USB port and Grove sensor interfaces. Otherwise, it’s almost identical to the BB Black. The Starter Kit includes 10 Grove gizmos, including temperature and sound sensors, relays, buzzers, accelerometers, and I2C hubs.

    DragonBoard 410c
  • DragonBoard 410c AWS IoT Starter Kit — Available from Arrow and Qualcomm, the DragonBoard 410c kit sells for $80, only $5 more than the standard $75 price of the companies’ underlying DragonBoard 410c SBC. No sensors are supplied with the kit, however.

    The maker-oriented DragonBoard 410c, which came in fifth place out of 53 in our 2015 Linux hacker SBC survey, is one of the first 64-bit ARM SBCs, and one of the first to use Linaro’s specification. The 85 x 54mm SBC integrates Qualcomm’s quad-core, Cortex-A53 Snapdragon 410 SoC clocked at 1.2GHz, as well as an Adreno 306 GPU. Other features include 1GB LPDDR3, 8GB eMMC, WiFi, Bluetooth, GPS, HDMI, a 13-megapixel camera interface, and USB ports. It offers the standard 40- and 60-pin expansion connectors.

    Intel Edison
  • Intel Edison and Grove AWS IoT Starter Kit — SeeedStudio’s $179 Intel Edison and Grove Kit builds upon the company’s Grove-augmented version of Intel’s $100 Edison Kit for Arduino. This in turn is based on the tiny, Intel Atom-based Edison COM board.

    The AWS configuration includes the Intel Edison for Arduino as well as a base shield and 11 Grove sensors. The board includes WiFi, Bluetooth LE, a microSD slot, and dual micro-USB ports. Other I/O includes a 70-pin connector and an Arduino breakout that supports Arduino shields.

The other seven AWS IoT Starter Kits use boards that run various RTOSes instead of Linux. These include:

  • MediaTek Linkit One and Grove — This Grove-enhanced kit is based on the MediaTek LinkIt board, which runs MediaTek’s new LinkIt RTOS in conjunction with an Aster (MT2502) SoC.
  • Avnet B4343W — Built around a Broadcom BCM4343W WiFi/Bluetooth SoC with Broadcom WICED SDK, this kit includes Avnet sensors.
  • Renesas — Based on a Renesas RX63N board, the Renesas kit runs on Micrium OS.
  • Marvel EasyConnet MW300 — This kit runs the Marvell EZ-connect SDK on a Cortex-M4-based Marvell MW302 module with built-in wireless.
  • Microchip — This kit is based on the Microchip WiFi Client Module, which uses a PIC32 microcontroller.
  • Seeeduino Cloud and Grove — Based on an Arduino-compatible Seeduino Cloud development board and Base Shield, this kit offers a variety of Grove sensors, DIP LEDs, and more.
  • TI LaunchPad — The LaunchPad builds on the Texas Instruments LaunchPad CC3200 board, which runs TI-RTOS.

“Arrow is excited to collaborate with Qualcomm Technologies and Amazon on this IoT Starter Kit,” stated Jes Lefcourt, general manager of IoT adoption at Arrow, speaking of the DragonBoard 410c kit. “We wanted to provide developers and makers with a low-cost, high-performance way to build IoT solutions — I can’t wait to see what kinds of innovations and venture-ready prototypes this kit will enable.”

Further information

All three of the Linux-ready AWS IoT Starter Kits are available for pre-order now, with shipments due later this year. More information may be found at the AWS IoT Starter Kit page, which offers links to all products, as well as Amazon’s AWS IoT announcement and AWS IoT website.

(advertise here)

Print Friendly, PDF & Email

Please comment here...