Physical hardware architecture from packaging to distribution

As some might know, I made a course change, career-wise, in the last quarter of 2025 and find myself in the field of distribution and fulfillment centers, after many years in the food packaging industry. One of the first discoveries is that, no matter the subject matter, the technology behind design seems to remain the same. Let’s look back at the features that make design of a packaging machine or packaging line easier.

Let’s start with the brains of the application, the processor. In my opinion, one of the greatest developments in recent years has been the merging of control and safety into one package. Safety PLCs have greatly reduced the hardware cost of a control system while making it easy to implement a properly designed safety circuit by taking out some of the magic behind the curtain.

A safety PLC is really two processors in one package. The safety side is usually distinguished visually by all the related devices presented in a red package, just like you would find on a separate safety relay and associated devices.

From a system architecture perspective, the hardware tree that tells the processor what is connected to it can assign safety and non-safety modules. Special instructions are used to monitor dual-channel inputs and create configurable redundant output instructions to perform the same functions as a physical safety relay.

Finally, for a safety PLC, the logic is executed in a separate task that is updated independently, meaning that the regular machine logic does not impact the safety functions.

Another feature that has made life easier for a designer is the integration of safety into variable-frequency and servo drives.

The first feature that comes to mind is the safe torque off (STO) function. This is the primary function in that the absence of the STO inputs will electronically disable the drives output pulses, removing the rotational magnetic field rendering the motor unable to generate torque. The motor will coast to a stop. This function can be directly tied to the output from the safety channels on the safety PLC.

Other safety features on the drive might include safe stop 1 (SS1). Once triggered, this causes a controlled deceleration of a motor and issues a STO once the monitored speed has reached zero. The STO alone will cause the motor to coast to a stop where this function causes a controlled stop before disabling the drive pulses.

If an encoder is used, safe limited speed (SLS), safe speed range (SSR) and safe direction features prevent unexpected speeds or unintended direction of rotation.

Combining the safety PLC with the safety features on a drive, one can utilize field-mounted I/O to create zones on equipment that is larger or on whole packaging lines.

One application I worked on used a safety PLC for line control. The PLC had dual-port Ethernet set up as two separate ports with one connecting the line control PLC to all the other PLCs in the room. This provided peer-to-peer communications between the various unit ops (machines). The second port was assigned at the machine level, connecting all the various local devices that controlled the conveyors and pneumatic devices on the line.

Get your subscription to Control Design’s daily newsletter.

The dual-port feature on a programmable controller has been a boost to design in and of itself. Traditionally, one would have one or more unmanaged or managed switches to connect all the machine-level Ethernet devices to a common network. If one wanted to have connectivity to a plant-level network, a managed switch or a network address translator (NAT) device was needed to map one or more local addresses to the plant network.

With the configurable dual-port Ethernet function, one can assign one port to the local network, for example, 192.168.1.10, to talk to all the machine-level devices and assign the second port to a host network, for example, 10.10.4.10. If all of the other unit ops are also on that same 10.10.4.x network, they can all talk to each other for interlocks and passing of parameters, without impacting the important local connectivity on the private, 192.168.1.x, network.

Unless your machine uses a lot of Ethernet devices that are constantly updating or are time-critical, one might forego using the more expensive managed Ethernet switch and go with an unmanaged switch at the machine level.

By using the dual-port technology described above, I was able to have all my major unit ops, 20 in total, on that host network while each of them had a separate private network for local function. The major unit ops were all sharing data, and, with that host network also having access to the next higher enterprise network, all our engineers could log in from their desks or from home and remotely support the entire line.

We assigned all of the PLCs and HMIs on that same host-level network so they could be monitored and edited remotely. This was a great boost to productivity, especially during off shifts when we traditionally would have had to wait for an engineer to arrive at work during normal office hours to support production issues that might arise.