Representative End User Clients
Representative Automation Clients
Representative Software Clients
This week I’m visiting some ARC clients in Japan as a follow-up to last week’s ARC Tokyo Forum. My main take-away from the Tokyo Forum (before jet-lag incapacitated me) is that you can’t have too many resources in an IIoT edge platform, especially if your installation needs to scale up or support mission-critical functions. Sadly, of course I’m being facetious. Edge equipment cost is part of TCO and higher-priced edge equipment will crimp an IIoT program budget.
But hearing Forum speakers expound about the need for low-touch or zero-touch deployment and support in the IIoT brought me back 7-8 years ago, to the early days of smart meters and the smart grid. Meters are a special case, but grid communications infrastructure was in its infancy then. Utilities with the deepest experience soon realized the great importance of speedy (meaning inexpensive) field deployment of such equipment. A utility that was installing say several hundred thousand smart meters would also be installing several thousand pieces of field support equipment as well. This was where the real “smarts” and communication capabilities were located.
These thousands of devices were installed by utility lineman crews or contractors. The products needed to be commissioned from a remote support center immediately as the crew completed the physical installation. From that point on, all support functions and updates (save hardware changes) needed to be performed remotely. This commissioning and support model fits the IIoT well. How closely do today’s installations follow it?
It’s the need for scale and growth that drives the demand for zero-touch deployment and support. Another resonant story comes from the early days of data center administration. I remember an early data center guru explaining his extreme need to eliminate manual operations with the administration processes. He told me that “manual steps that are acceptable today will be a pain point when the operation grows by 10x and will become intolerable when it grows by 100x". Data Center automation requirements were high because scale-up was rapid and one couldn’t tolerate linear growth in administrative labor.
What I take away from this is the need for end users to compare the costs of various installation and support options carefully at the very front end of an IIoT project – even though the various options and their costs will have much more uncertainty at that point. This need does not disappear if the project is a “product as a service”. Instead the analysis requirements shift from the end user to the supplier, who will need an effective deployment and support strategy to manage his own TCO as the installed base grows.
But hearing Forum speakers expound about the need for low-touch or zero-touch deployment and support in the IIoT brought me back 7-8 years ago, to the early days of smart meters and the smart grid. Meters are a special case, but grid communications infrastructure was in its infancy then. Utilities with the deepest experience soon realized the great importance of speedy (meaning inexpensive) field deployment of such equipment. A utility that was installing say several hundred thousand smart meters would also be installing several thousand pieces of field support equipment as well. This was where the real “smarts” and communication capabilities were located.
These thousands of devices were installed by utility lineman crews or contractors. The products needed to be commissioned from a remote support center immediately as the crew completed the physical installation. From that point on, all support functions and updates (save hardware changes) needed to be performed remotely. This commissioning and support model fits the IIoT well. How closely do today’s installations follow it?
It’s the need for scale and growth that drives the demand for zero-touch deployment and support. Another resonant story comes from the early days of data center administration. I remember an early data center guru explaining his extreme need to eliminate manual operations with the administration processes. He told me that “manual steps that are acceptable today will be a pain point when the operation grows by 10x and will become intolerable when it grows by 100x". Data Center automation requirements were high because scale-up was rapid and one couldn’t tolerate linear growth in administrative labor.
What I take away from this is the need for end users to compare the costs of various installation and support options carefully at the very front end of an IIoT project – even though the various options and their costs will have much more uncertainty at that point. This need does not disappear if the project is a “product as a service”. Instead the analysis requirements shift from the end user to the supplier, who will need an effective deployment and support strategy to manage his own TCO as the installed base grows.