IoT @AWS re:Invent 2021: The Hotel California Effect

IoT @AWS re:Invent 2021: The Hotel California Effect

author portrait

Written by Gaurav Suman

Category: HiveMQ IoT

Published: December 15, 2021


10 years of AWS re:Invent meant AWS was going to make this one matter. And they did.

I am sure you have heard quite a lot already from AWS about the event, and the commentary that has followed. I am going to cut through that clutter, and share with you my perspective which - I hope - will be useful to people in the IoT. Here we go.

Managed and Packaged Services took centre stage

When Adam Selipsky took over the reins from Andy Jassy, he was clear that AWS will build more products for specific industries. At re:Invent this year, we could see that coming to life.

With AWS Fleetwise, they want to make it easy for a business to manage vehicles and people on the move. This includes on-board devices, code, dashboards, etc. One might think this applies to only logistics companies, but think about the use cases for rental cars, ride hailing, food delivery, etc. They all have people or cargo on the move and being able to track and manage for predictability, safety and billing purposes, are important use cases.

On the same theme (uncannily) is the new service Roborunner. This is fleet management, but for robots in a facility. As robots continue to take hold on the factory floor and elsewhere, their management and interoperability is getting more complex. This is especially relevant when we require different robots to integrate and orchestrate tasks. Roborunner aims at solving that problem.

Then there is the AWS IoT Twinmaker. Its raison d’être is “easily creating digital twins of real-world systems”. The idea is to help you create a 3D view of your systems and operations to optimize efficiency, increase production, and improve performance. As you can imagine, each virtual entity is replicating your real physical systems, and there is a great opportunity to model and back test scenarios.

AWS is coming for the edge and the endpoints

Some industry observers point out that with their announcement of Private 5G, AWS is eating the telcos’ lunch. For a couple of years now, several telcos have partnered and built services on the back of AWS Wavelength. For the uninitiated, AWS Wavelength is a physical server (or servers) you can install in your data center and virtually extend the AWS cloud into your data centre.

Those who follow AWS closely will know that they have, through their Sidewalk project, already made clear their intent of cheaply connecting devices to the internet. The keyword there is “cheap”. Projects like Sidewalk and products like Private 5G get them closer to disintermediating the telco for the end customers.

Also, in 2021, their focus has moved closer to the endpoints and sensors. AWS - with IoT Greengrass and their partners - already had good visibility in this space and they have now worked with module makers like uBlox to create custom wireless modules that hook up to the AWS cloud by default and immensely reduce the value of the code on the device. These are now being built to meet the IoT Expresslink specifications.

A bite of reality

Remember this timeless line from a popular Eagles number - “You can check out anytime you like, but you can never leave

That’s a factor here. Let’s call it the Hotel California effect.

Now, what’s not to like about all the work that AWS is doing on your behalf ? Nothing… really.

They help programmatically automate your fleets, get a low-down on your smart building designs and frameworks, and abstract you away from your whole value chain - sensor to the dashboard. The question is, are you sacrificing something that is, or can be, your competitive advantage?

If yes, then passing off key stages of the value chain can hurt in the long run. Imagine wanting to strategically source and stream data from your sensors to multiple destinations. To AWS for storage in S3 and to Azure for data mining in Azure ML. Are you setting yourself up for that flexibility?

A prudent architectural choice will be to use an open-standards based IoT broker like HiveMQ on the edge, and that will help you keep your options open on where to analyze, where to store and where to take action.

The choice is between convenience and freedom. Businesses and vendors alike have expressed serious concerns with hyperscalers like AWS where it’s easy to bring your data and IP in, but hard to migrate or expand to other platforms. The Hotel California effect.

AWS is pushing the boundaries for what can happen with the power of the Cloud. Businesses have a lot to gain from it, as long as they are strategic about what they choose to use and how.

If you are starting out on your IoT journey and want to benefit from HiveMQ’s perspective of helping 130+ IoT customers worldwide, reach out to us for a free consultation.

About Gaurav Suman

Gaurav is the Director of Product Marketing at HiveMQ. He is an engineer by education, and is passionate about helping businesses thrive with the help of technology.

Contact Gaurav

newer posts Part 3: Sending sensor data from Raspberry Pi Pico to HiveMQ Cloud
Event Driven Microservices Architecture for IoT Solutions Using MQTT older posts