I hate the term IoT – the Internet of Things. It is just wrong on so many levels. What most people fail to realize is that IoT is, like most things, about more than technology. But whenever someone starts mentioning IoT, the acronyms start flying around; MQTT, AMQP, DMP, OPC UA, IoT Hubs, and then Microsoft Azure, Amazon AWS, and Oracle IoT Hub. Vendor are forcing their version of IOT. Why can’t I have it my way?
Everybody wants to be in the IoT business now. Of course they do. There’s going to be a gazillion nodes of IoT in the world. Everything will be wired. When you get in bed, your bed will check with your toothbrush to make sure you brushed your teeth before bed and the bed won’t let you sleep until you brush your teeth. The egg carton in the refrigerator will nag you about buying eggs. Your nail clippers will remind you it’s time to cut your toe nails.
There is a whole lot of silliness in the IoT business
First, let’s start with the business model. Everyone wants to collect data, lock it up in their proprietary server, and charge the customer $9.95, $49.95, or $199.95 a month for the service. I had one guy quote me $79 a node to collect data from an industrial node. Who is going to do that? Who wants to have their data held hostage by some provider that you’re not even sure is going to be in business next year?
Then there’s the technology end of this. I see people building their own hardware to do this stuff. I keep wondering if I am missing something here. Let me state it plainly: “HARDWARE IS A COMMODITY.” I recently had a talk with a very large IT company that has always been a hardware company. They’ve built a really nice, super rugged box with oodles of memory and features for IoT applications. I gently mentioned to them that I can get Taiwanese hardware to do those applications for 10% of the price. Yes, 1/10 of the price!
The software is relatively straightforward too. Everybody and his sister-in-law can implement an MQTT transmitter that sends data on topics to a broker. That’s hardly an accomplishment now. But the part they are always missing is the application layer. What’s in the message? How is the receiver going to decode it? In the applications I’ve seen, there has to be prior knowledge coming from some non-digital description of how to decode the message packet contents. That’s not a good way to do it.
At RTA we’re doing things more simply. As you all know, I am a very simple guy. I like pizza, football, and wine, and I hate complexity. At RTA, we’re going to build some IoT devices, but only to move factory floor data into the enterprise. And we’re going to put them in your hands, Mr. Customer, so that you can use them, control them, and manipulate them at will.
If you want your manufacturing data in a SQL database or an Oracle database, we’ll put it there for you. If you just want the data in Excel we can help. If you want it in AWS or MS Azure or Oracle’s IoT hub, we’ll get it there for you too. Or if you want it in your own cloud, that’s OK. I see it just like the old Burger King commercials: Have it your way.
This is nothing more than what’s been going on in manufacturing since day one. From the time of the first PLC and then EtherNet/IP, ProfiNet IO and Modbus TCP, customers have wanted to collect production data from the factory floor and integrate it with business systems. There is nothing new today. There’s just a lot more data to collect and more things to do with it.
And RTA plans to continue helping customers do that – let’s just not call it IoT.