The term “Industrial Ethernet” can mean so many different things. And the sandbox just got a little bit more crowded.
I have written about time-sensitive-networking (TSN) Ethernet, which is supposed to provide the user and the process with data transmission determinism.
Should TSN Ethernet have been available 15 years ago, my implementation of a maintenance data network at a large distribution center would have been a lot cleaner.
The real problem we had was the interface with a custom-written warehouse-management system (WMS), which connected to a network of Rockwell Automation’s legacy Allen-Bradley PLC-5 Ethernet controllers.
An incoming pallet would have its barcode scanned, and that information would be transferred to a serial terminal server that was connected to the UNIX-driven WMS. This barcode was then transferred to the PLC-5 responsible for that pallet location, of which there were three.
The barcode was placed in a database on a maintenance server for data logging, using a custom Microsoft Visual Studio program. Once the destination of the incoming pallet was determined by the WMS, that information was again transferred to the PLC and then again to the data-logging service.
The PLC then processed a release command from the WMS upon confirmation of received data, and the pallet went on its way. There was a “pallet arrived” message that was created by the PLC to the WMS so that the WMS database could be updated.
Needless to say, there were a lot of moving parts, which created some really funky timing issues with data movement.
I was not privy to the WMS source code, but suffice it to say that if the messaging got delayed in any way there was a distinct possibility that the content of the messaging was misinterpreted and the data got applied to the wrong pallet, creating a nightmare for inventory.
If TSN was employed, I suspect that we would have had fewer data-transfer issues, so this is something that you may want to look into for your future projects that are time-sensitive.
There are two new specifications that have made their way into the networking sandbox. These are Ethernet-advanced physical layer (APL) and single-pair Ethernet (SPE).
Ethernet-APL is an intrinsically safe two-wire Ethernet physical layer implementation. Being intrinsically safe will allow this technology to be used in hazardous-location applications.
Single-pair Ethernet is the non-intrinsically safe version of Ethernet-APL.
So, what could be the benefits of these new technologies? And why will they matter?
The world has been built on 4-20 mA twisted pair forever, but with the advent of Ethernet and high-level protocols the amount of data that one can get from a configured sensor is enormous. This amount of data needs a big pipe to transmit to the mother ship, and a current loop may not be the physical layer of choice.
I asked a colleague of mine who is involved in the market of instrumentation about the percentage of Ethernet-enabled sensors versus 4-20 mA, and I was told that 1% of sales over the past 20 years was Ethernet. You guess what my reaction was.
One of the benefits of the current loop is cable length, which can extend upwards of a few miles. Standard Ethernet is 100 meters, so we have an issue.
Ethernet-APL however has a cable length of 1,000 meters at 10MBit/s. In a standard plant that would cover most if not all current-loop applications. And therein lies the primary benefit of SPE/Ethernet-APL. It can use existing twisted-pair wiring up to 1,000 m.
So, you can replace a standard, legacy current-loop flow sensor with a modern-day Ethernet-enabled protocol-based device using the same wiring.
You will need SPE/APL switches and routers to build out your new network, however. Many vendors are offering these upgrades presently.
Also read: Technology gets a facelift
When I was talking to my colleague about Ethernet-APL, his first comment was “It is real.”
I also contacted the guru of fieldbus, Ian Verhappen, a fellow Canadian. As luck would happen, he had just done a presentation on Ethernet-APL to the International Society of Automation (ISA) chapter in his area.
I learned a lot from his presentation, such as the fact that Ethernet-APL such as it supports power as well as protocol, increased spur length distance over standard Ethernet, supports multiple protocols and has an unlimited data payload.
The IIoT market is expanding at a high rate of speed, as are edge-connected devices, which can be such things that enable sensors. Verhappen suggests that most IIoT devices consume less than 10 MB per month, so the speed of Ethernet-APL shouldn’t be an issue. He is waiting for that killer app that explodes Ethernet-APL usage in the industrial space.
There is an organization supporting Ethernet-APL, as well. So yes Virginia, it is real.
