We are going back to the mainframe model. The clients are anything, the server is the cloud and the method is the Internet. Software as a Service is redefined. Your car will be a client.
The cloud is a representation that defines networks. It is somewhere out there. Servers, virtual or not, hold software that we can access on per-use terms—monthly, subscription contract, or corporate license.
Office 365 is a new Microsoft service that, for six bucks a month, lets users access office applications. Microsoft says you can access any of "your" data from any device, so you can be anywhere and edit and print any document from the server over the network. Google has had this for a while, but it seems that this new model is being embraced by more.
There are advantages to working in the cloud. You still have a single point of blame, but it's not in the building. It can decrease risk depending on how you define risk. Your data is stored on the server in the Land of Oz. Is it secure?
The network side of it is probably less risky. Alternate network paths and providers give as close to 100% uptime as possible. Maintenance of the office goes away. Virus and spam management is on the shoulders of the cloud provider. Maybe the IT guys can do more productive things now.
Amazon has a cool service called Elastic Compute Cloud (EC2). It allows users to rent resources in an off-site data center(s). Salesforce.com is and has been very successful providing CRM-based cloud services. Route1.com says it can provide true secure access to the cloud and VDI applications if high-level security is an issue. Ortsbo.com offers real-time language translation.
Why is this important to a machine builder? Things seem to be in place for something really big. But what might that be?
We build machines, and processes, and job-specific applications. The end user usually isn't anywhere that we are. They are global. They don't speak our language, and most of them don't care about intellectual property.
Can we provide a machine that can be serviced only by "us"? What about something that takes away the need and, in fact, the flexibility to be changed by the customer? We can't reprogram our own devices. Why can't a machine tool and/or process be like COTS software?
You could monitor and alter things over the Internet, securely, and communicate with anyone in any language. You could do this from any location in the world that has access to the cloud.
It would be cost-effective and promote business expansion by adding value to your processes. Your machine could call home when it is in need of service—predictably. Software upgrades might be an option because you would have a global application suite that is cloud-based. Once we've figured out how to successfully connect our processes to the cloud, then we could grow that part of our business the way we want.
If you can virtualize a server, why can't you virtualize a PLC or control system? Maybe not, but the interface is a guarantee. There has to be more than a web server in the hardware, though.
I asked a local machine builder about legacy upgrades, and the ability to globally monitor machines like Maytag does. "Not interested," he said. "We build 'em, and ship 'em."
Colleges and universities provide training virtually. Costs are rising, and inflation is coming. Why do we keep making the same mistakes and doing the same things the same way? Change can come only from the necessity of a crisis, which I believe is on its way. China is expected to become the biggest economy in 2016.
Probably nothing can stop that, but they will need our IP to feed it, and our manufacturing expertise. Allow them to build, and place. We need to own the brain cells that put it together.