Google Cloud IoT Core Retirement Thoughts

This recent news has reminded us of a common situation organizations find themselves in when building out their connected solution offering. Many companies have chosen to build a solution on one of the large cloud providers because of the marketing of IoT with promises of scale and big data while feeling ‘safe’ that they were partnering with one of these big companies.

Quick thoughts

Having to re-provision hardware edge IoT devices is non-trivial and for those caught in this situation hopefully you have a good OTA system to update your edge firmware / certificates for your devices.

IoT is more nuanced than just providing an API or a webservice. You have lots of unique edge device sources providing data in different formats and different report rates.

Regarding this decision, most likely:

1. Google is evaluating the most profitable areas of the business and eliminating others to be able to focus.
2. Google likely decided that rather than supporting customers with all of these IoT details and complexities, they could instead have customers manage it elsewhere but still use and pay for higher profit generating services that require less hands-on support.

Google is assuming IoT device data is still going to make its way in but as a unified stream of data or a dump into their other services for things like analytics and machine learning (ML). The IoT part just requires too much hands-on support, especially for small and medium sized companies.

Cloud IoT services history

Backing up several years ago, AWS, Microsoft Azure, and Google Cloud began offering ‘IoT’ services, which many believed would be used by the majority of organizations looking to create a connected solution and digital strategy around machines, equipment, and services. Their main motivation was not to offer comprehensive IoT services, but just provide a device ingestion layer to get your data to their more expensive services.

Exosite had already been running a platform for several years when this happened, with years of IoT experience. It was of some concern at the time, but also was a point of validation of the need for IoT platforms with off-the-shelf solutions. Our platform had specifically been created to handle a high volume of different types of IoT device connections and streams of data from sensors while providing a user interface for monitoring of these high value assets.

Digital success for everyone?

With the entrance of the cloud service providers providing an IoT layer, it was assumed that this would be the way every connected project would go and it would simplify the buildout of solutions. What you may not hear about though are the case studies of organizations that were not successful.

Here are some observations:

1. Most organizations who build machines and offer maintenance/support service for equipment have little to no internal experience for building with cloud services, which is required in these situations.
2. An internal team member can use tutorials and can create a prototype to show a few connected devices talking with some basic interface to see the data. It’s then assumed it’ll work for 1,000 connected assets and then 10,000 but it unfortunately isn’t that straightforward.
3. Every application of a connected asset is different, even multiple types of use-cases within the same organization.
4. The allure of ‘partnering’ with the big tech companies is there and it is hard not to get excited about.
5. It feels that it’s a safe bet.

Creating those big system diagram charts in your IT department with all of the boxes of the different services you are going to use from the cloud vendor can be exciting but what can be missed is that your organization now owns defining, connecting, managing, scaling, and operations of all of it. IoT involves not just a different skill set, but several skill sets from end to end. It’s just not traditional IT.

Reality

Let’s all be honest. The big providers want to sell cloud server time. They do not sell solutions, having an IoT component service is a way to sell more server time by ingesting more data.

You almost always have to work with a certified partner with the cloud vendors, which means you are paying someone to build something customized for you and likely also paying someone to be a consultant to you on how to build it.

You may not even know if your digital business model strategy is legit but it could take well over a year to have something to prove it out.

What about Exosite’s IoT platform?

You might ask, “Does Exosite use cloud IoT service?” The answer is No. While Exosite uses cloud computing to deliver our services, we do not rely on any cloud provider’s IoT services. We built our own IoT connectivity layer to ensure we could provide the functionality needed by our customers, ensure we could scale it, and ensure we have ownership. In fact the original Device APIs have been supported now for over 10 years, while we have added additional protocol support and features.

Summary

The two takeaways hopefully here are the following:

1. Depending on your organization size and team, taking on a project building out your connected solution on cloud services may not make sense - there are ways to get a scalable ‘cloud’ IoT solution to prove out your business model and a solution in weeks vs years.
2. The large cloud service providers do not offer IoT Solutions. They provide web services and cloud computing, not a ready to go solution. Connecting different types of edge devices requires too many details per the application use case, they are just too big to be able to support you well.

If you happen to use Google Cloud’s IoT Core for your connected offering and need to find another solution, reach out to Exosite’s team. We can walk you through what it would take to update all of your IoT device endpoints to send data to our platform (we offer both HTTP and MQTT device APIs, noting we have supported the same HTTP Device API for over 10 years). We can also discuss leveraging our off the shelf condition monitoring application ExoSense. It may not be a fit and that’s ok, but we will work to figure that out quickly so you can find the right path forward.