Google just announced the phase-out of its Cloud IoT service. If you just read these headlines then you might be missing out on what this means to you, as a manufacturer. So, what does it mean?
In my opinion, this indicates that one of the world’s technology leaders could not make a commercially viable IoT hosting platform. I’m intentionally saying “commercially viable” because Google has the talent and cash to do just about anything. For proof of this assertion look no further than their AI, which is so advanced that one of their own engineers made waves by claiming it was sentient (read the transcript, it’s frighteningly impressive). No manufacturer that I’ve encountered (which is more than a few) has aspirations of copying Google, so what does this mean to you?
The most important takeaway should be comforting, if not anticlimactic: the difference this makes to Digital Manufacturing will be imperceptible. Why? Google’s use case is not a manufacturer’s use case. This may seem like an obvious statement, but enough executives read the headline “Google Stopping Cloud IoT” so it’s worth noting what makes your IoT adventures different from Google’s. They care(d) about IoT, and you care about IIoT: emphasis on Industrial Internet of Things, in this case. Take their communication protocol decisions for example, support for MQTT and HTTP but no mention of OPC UA. This omission means their target was never manufacturers, at least to the extent that their product alone was not meant to be a cornerstone of your digital manufacturing efforts. Sure, their product suite was used by companies trying to provide Cloud SaaS solutions with Google’s IoT, ML and AI. But manufacturers would have interacted with those companies, which often leverage Microsoft and Amazon’s services too.
Though Google’s decision doesn’t impede Digital Manufacturing, there are lessons to be learned:
Lesson 1: It Confirms that Applications Are Still Important
An ankle-deep wade into Google’s Cloud IoT page reveals a host of other solution names. From their website: “Use downstream analytic systems by integrating with Google Big data analytics and ML services such as Cloud Dataflow, BigQuery, Cloud Bigtable, ML, Google Data Studio.” Did you see any announcements about shuttering its Google Cloud Platform or any of their data analytics tools? Neither did I. This is an editorial leap, but I don’t think that Google’s IoT had a solid application. It was a side project meant to drive Cloud revenue, and it missed the targets.
If Google can be humbled like the rest of us, then the adage about lacking plans and getting nowhere remains unchallenged. A survey of manufacturing outlooks from the last few years will conclude that well-designed lighthouse plants are seeing significant gains, while plenty of organizations are still struggling with Industry 4.0. This is not a paradox. There are many things that separate the winners from the losers; starting with a measurable application in mind is a consistent success factor. That’s not to say that new and worthwhile revelations are a hoax, but they are typically reserved for organizations who observe the periphery impacts of their initial success.
Lesson 2: It Informs Us on How to Approach IoT-enabled Products
I started by making the case that Google’s Cloud IoT platform was, on its own, not a good fit for industrial purposes. Therefore, its phase-out doesn’t predict doom for Digital Manufacturing. It might, however, shift things in the realm they are intended to be in: consumer-level IoT and smart products. Even if you don’t manufacture smart products today, it may not be far off. While we’re evaluating a platform misstep, let’s point out a few other areas where IoT can fall short.
Poorly Designed Smart Products Are Dumb
Deciding what you manufacture comes before deciding how you manufacture it. IoT products are popular but they aren’t always logical. Here’s an example from my life: as a parent of young children, my recent consumer tech expenses have been centered around kid health and safety. Smart heart rate monitor socks, smart thermometers, smart A/V monitors, smart sound machines, and so on. Some of these products have enhanced my experience. Others just haven’t. We bought a fancy sound machine that could be programmed through our phones: schedule, volume, bird tweets vs ocean waves, etc. The concept was great, but the IoT platform execution was so poor that I now prefer inserting and removing the power cable to turn it on and off. Ironically, the “smart” machine is far more cumbersome than our other kid’s white noise generator that’s essentially a fan with an on/off toggle. Adapting a quote from one of my favorite doomed characters, Qui-Gon Jinn, “the ability to connect does not make you intelligent”. If you can, avoid digitizing just to say you’ve done it.
What Can You Realistically Influence?
Let’s say you have a good case for a smart product. Now you have to make it.
Start by getting involved early. Your organization may employ “design for manufacture” (DFM) or a similar method that involves operations in product ideation. You can be a voice of reason in this process by ensuring that the “smartness” of your product has a real application. When the design moves forward, consider how your manufacturing process will be impacted by including new electronics. Producing IoT devices itself may open new challenges and opportunities in Digital Manufacturing. Think back to the two sound machines — the “smart” one required circuit boards which were probably serialized. Including the PCB doesn’t just add an assembly step, it increases your data collection and traceability. Are you ready for that?
Second, make a holistic plan. This includes ensuring your information systems will be enablers instead of blockers. Do you have a quality management system that facilitates Advanced Product Quality Planning (APQP) and Failure Modes and Effects Analysis (FMEA)? Is your shop floor system ready to check skills and display digital documentation while supporting Industrial IoT and operator-entered information? Google’s business model had little to do with preparing organizations for the introduction of IoT. I can’t claim that this caused their downfall, but we should approach the correlation with cautious respect.
To conclude, I think it would be a mistake to let Google’s IoT phase out negatively influence a manufacturer’s opinion on the value of IoT, especially IIoT. I also think that it would be a mistake to completely ignore the announcement. Achieving success with (I)IoT and Digital Manufacturing requires a focus on application and attention to execution. Keep moving forward, and keep your heads up.
