Sept. 11, 2008 – We spoke with Raj Saksena, CEO of Omnitrol, about their recent recognition for innovation in plant automation.

Congratulations! Omnitrol won the M2M Award (Machine to Machine technology). What was the name of the partner that you worked with on that?

SAKSENA: The customer was Endway Defense Systems, a manufacturer of high-end microwave technologies for fighter aircraft.

When you first approached the project, how was the goal determined and how was it expressed?

SAKSENA: The company needed more visibility and traceability, to being able to audit various manufactured parts; being able to track, in a reasonable amount of detail, the status of what is going on with any manufactured parts; and ultimately to get them delivered to their end customers.

What have they been doing up until this point before the new technology?

SAKSENA: Most of it was manual. They were doing a lot of steps where information has been gathered by logging it through terminals or through barcode, scanners. At least there were something, but definitely not sufficient to give you a complete traceability and/or visibility to all aspects of the manufacturing process.

When thinking "Machine to Machine," I am assuming that, of course, the goal was to have two machines be in communication with each other, with no human in-between. What types of machines were these?

SAKSENA: Mostly they were a set of RFID scanners and readers, mobile devices, which were hand held, both scanners and input devices on the shop floor. There were ELC controllers that drove various outputs that give indicators or alarms and alerts to people working on the shop floor.

We had this sort of floor loop where we are taking input from the sensor network and delivering output that controls the system that provided feedback to systems that were used during the work process.

In hand held devices, the first few that comes to mind is if I am holding a device and you are holding a device, I might call, “Hey, Raj, stop the machine!” But, in this case, is it the device that is talking to each other or are they talking, not to another handheld, but to some equipment on the floor.

SAKSENA: This is not really information that was being gathered from another person working on the floor, but sometimes that information would be correlated in the working process, which was now made visible to someone who was either querying it or updating their work order, or it would give information on what to print, which was now available in the handheld and they could execute the task at any location.

What was the Omnitrol technology that was involved?

SAKSENA: From the point of view of an architecture, Omnitrol built the integration platform. It integrates the sensors, the PLCs, information retrieved from the sensors, the integration of a mobile device, therefore, all the triggers that were coming in from a mobile device, for example, that would result in it printing of, let's say, some information for the product.

The Omnitrol would take the information and then print it out on an RFID printer. We eventually had the Omnitrol acting as the automation between the way in which you are taking data input or input from the RFID reader or other sensors and then providing the output, which is going into another device or to a system.

Was there routing involved in this where perhaps some information went to one particular machine and different information went to another machine?

SAKSENA: Yes, absolutely.

For everything it is very contextual. The business process that we provide visibility into eventually means that you are providing some context of what the event is.

When you put that context in perspective, you may sometimes drive output that goes to what we call a smart shelf, or a smart bench, or it goes into the inventory systems or to the inventory people, to receiving or shipping and so forth.

Sometimes input coming in from a business context standpoint will result in the output being visible only to that aspect of the business process that needs to know that event occurred.

Are each of the devices communicating somehow through either a central controller, or central system, or is each one more like a node that can communicate with other nodes?

SAKSENA: The Omnitrol is the integration layer, which means it includes all the networking and the switching between these elements; so, Ethernet based devices, Wi-Fi devices, aerial port, multiple drop devices, motion sensors, humidity sensors, different types of devices feeding this, are actually all physically wired back into Omnitrol, which is sitting on the rack.

All the application logic that deals with the processing of the event and the end result, which may be a display on a hand held changing or an event going back down a smart shelf or to some other element on the shop floor, are applications running inside the Omnitrol server appliance. What you essentially have got is an intelligent sensor network, which is now has got business contexts as part of the way you execute services, or residing essentially in this one integrated factor.

You mentioned Ethernet. Were there different types of input sources and wave lengths involved, or was it all done wirelessly on one particular frequency?

SAKSENA: No, we had multiples. We had interfaces coming in over Ethernet, over WI-FI. We had interface coming over for PLC control, over multiple drop interfaces. So you have this input sensor, motion sensor, fact lights that are controlled by a PLC, which eventually have serial interfaces to them.

Were the sensors you mentioned just now examples of what could be used, or are those actually used in this case?

SAKSENA: No, they were actually used in this deployment.

So they have gone from more of a manual system to something that actually not only had machine to machine communication, but had sensors contributing their two cents, if you will? They went from manual to all these different devices?

SAKSENA: That is correct. Basically, the end result of a deployment like this is you get complete visibility of what is going on the shop floor, with absolutely no manual intervention. There is nobody logging, nobody stopping to do something, to log what the status of their working process.

The way we view it, it is almost like the shop floor is talking to you. So while the people working on the shop floor are doing their jobs, all the traceability and visibility information from a business process context, is going directly to management.

Is this too soon to have any metrics in terms of either man hours saved, or dollars saved or have they been able to determine that yet?

SAKSENA: No, we provide all of those metrics. The analytic runs in conjunction with the information being gathered, including all timing information, delays, which has worked for how much time on a given task. You can eventually analyze quickly, in real time where you may have a problem -- how much work was brought back in for rework, which means it failed QA.

You can deal with revenue per day. You can track the products you are manufacturing and get a trend, which gets you down to the financial metrics as to how well are you operating from a revenue standpoint, number of products manufactured per day and so on and so forth.