Industrial Automation Q&A: In-House vs Contractor, ET 200SP Retrofits, and Learning PLCs

How's it going, everyone? Vlad here. Welcome back to this week's episode where we discuss different questions and different comments that we have coming in from our industry. We have five very interesting questions that we will be talking about. So without any further delay, let's get into it.

And so the first question this week comes from the username Froggie Mayo, and he says, "My office desk for the week, spinning yarn machine conversion, Siemens ET 200SP plus Odot I/O." There are a couple of interesting images. So here is his laptop on a spool of what seems to be cable. This does not seem to be a spool of yarn that we maybe see in the back. We have a very old panel, just a lot of different contactors, a ton of fuse blocks, some relays, and a lot of different I/O. I cannot even see where the PLC is.

We have the actual machine. Again, it is a little surprising that he is posting this publicly. I do not believe that there is any intellectual property, but I have personally always been told never to post any customer information. So we have the machine that the customer is utilizing for spinning yarn. We have the new panel.

I am assuming that this is the new panel. This is the ET 200SP I/O adapter, as I am assuming, and that is where we have the Odot terminal blocks or I/O blocks, which all land inside of this terminal strip where you can connect I/O. It does not seem like there is anything connected on the opposite side of these terminals, so either there are maybe two terminals on the top and that is how they have connected to the field, but more likely there needs to be something that comes out of these blocks once the machine is commissioned.

And he says, "Hello, just wanted to share my setup and current project keeping me busy on site this week. Nothing quite beats the classic cable spool desk. When you are out on the floor, you can check out my highly ergonomic workstation. The project, I am doing a complete control conversion on a massive spinning yarn machine. The old hardware was well overdue for a modern upgrade. Before, we all know and love the classic spaghetti monster cabinets that come with these old textile machines. This one was definitely showing its age and made troubleshooting a nightmare to see what I was dealing with initially. The upgrade ripped out the old gear and migrated everything over to a new Siemens ET 200SP system."

He did not share what kind of a PLC he is using. Maybe he left the old PLC and this is just a remote I/O panel upgrade. Maybe he did move away from a different controller onto a more modern one, either an S7-1200 or S7-1500 series. "To handle the massive amount of I/O, we are using Odot digital input and digital output modules."

This is very interesting. We will come back to that in just a moment. "You can see the new clean panel coming together. It is incredibly satisfying to go from decades of tangled wires to neatly tacked and labeled blue wiring. I will check back in and update you in about two weeks with a full breakdown and a video of the machine completely up and running. Happy commissioning, everyone."

So a couple of thoughts. Number one, like I said, he is not necessarily explaining which control system this is coming from. He is utilizing the ET 200SP I/O series from Siemens. This is very common for these types of applications for remote machinery. Maybe he has a PLC that is controlling multiple pieces of equipment. That is also very plausible, and he is using the Odot I/O, which is surprising. I would assume that you can pick the same catalog numbers out of the ET 200SP modules. There might be some cost savings. I have not personally used the Odot hardware. It comes from China. There is documentation and there are manuals that you can download from their website.

I was not overly impressed with just the quality of materials. Again, I am assuming that once you have figured out how to properly integrate it into whichever control system, it probably works just fine. I cannot imagine that the savings are that significant, but again, maybe this is a customer that is very price sensitive, so applause to him for making sure that is handled.

In terms of going back maybe to the first panel, I did notice that there does not seem to be a controller. So maybe a lot of the I/O, and we can also tell that this is maybe a different panel. This is a green panel as opposed to what they have put in place. So I do not know if they replaced the entire cabinet. There does not seem to be a controller here, and even if I zoom in, I can obviously download the image and try to figure this out. There does not seem to be an actual controller, so maybe a lot of the wiring is coming from an auxiliary cabinet. We have the actual equipment. Oh, I stand corrected. It is the same green panel as we have seen in the past image.

So it is remote I/O in this instance. I would assume it connects via PROFINET to the main cabinet with the controller, as, once again, we do not have the controller. It looks like we have also cleaned up some of the other contacts that were already in place in the previous system. We have the contactors at the bottom. We have a couple of fuse blocks. We have the relays. We have the large fuse blocks and disconnects. So this is a job well done. We will see what it will take or if there are any challenges that he describes as he, quote unquote, "updates us in a couple of weeks." And hopefully there will be a little bit more information on the system, as I have not personally worked on textile. So this is definitely of interest and of curiosity, I would say, on my side.

All right, and the second question or comment is going to be incredibly interesting. I do have multiple stories around this specific comment. So we are going to switch over and read the username's post. The username is The New Wave Order, and he says, "Customer said he loves to fix on his own. Now it is not working."

And we have an image of a panel that is attached to the post. We can very quickly tell this is an older Siemens controller. We will see the RS232 protocol that supposedly, or I would assume, goes to an HMI. We can then notice multiple relays of different nature. A lot of them are twenty-four volts, which is a good sign. We have some Schneider Electric relays as well sprinkled in, meaning that something has definitely been replaced over time. We have a couple of fused outputs.

And so we have that tied to the field equipment. Overall, this is not a very complex panel. So if you were to reverse engineer something like this and migrate it to a new platform, probably a couple of weeks would be sufficient. We have the in-feed voltage coming in. We have some strange wiring. So wire nuts usually indicate the North American presence. So I would assume that this is a machine in North America, because in Europe you will see a different style of clamps, even inside of equipment, but also in the residential space. It looks like we do have some drawings on the back, so maybe there are some hints or perhaps some tips as to what this looks like.

And so we do not have anything else besides the image and this brief comment. And so as I have mentioned, I do have a couple of stories from the industry. A lot of times as an automation systems engineer or control systems engineer, you get brought in to solve some of the problems that the customer was simply unable to resolve on his own. So typically that means they probably have electricians on staff, they probably have engineers on staff that would be doing routine maintenance, perhaps some updates to the panels, perhaps upgrades and migrations from an old system to a newer one. But the reality is there is often simply not enough work for them to do, so there are not going to be large teams working on these unless you are working with a very large customer.

At which point they normally bring you in. Of course, they would have preferred to have done it in-house as opposed to bringing in an external consultant or a contractor. That, number one, usually is going to be more expensive, and ultimately is going to know the process or what is going on at the plant level a little bit less.

And of course, there are pluses and minuses in utilizing either resource, so there is going to be a long conversation to be had on whether or not you want to assign your internal electricians and engineers to something like this, or if you would like to bring in an external contractor. On one side, you have resources that are typically employees that need to be paid, that have already assigned work for them. On the flip side, you have a contractor that can be, in theory, fired depending on the clauses. I have seen a whole lot of different situations in my years. But ultimately, there will be a time where you can just let them go because there is simply not enough work to work on these types of upgrades or fixes or what have you.

Now, I have also seen many different customers that will decide to go the route of retrofitting a machine or fixing up a machine on their own. They will then fail for many months, sometimes years. And then they will bring in that resource, not realizing how much time, how much effort, and ultimately how much was put in to try and fix this in-house, which would have otherwise been extremely inexpensive comparatively speaking to using a contractor or a consultant.

So what I see a lot of, I would say, inexperienced or younger maybe engineering managers and directors miss is the fact that they are allocating human resources and capital first and foremost. So when you start to create a business case, whether we are going to do this in-house or whether we are going to bring in a contractor, you need to weigh in all the pros and cons, how long it is going to take, how much it is going to cost, what the hire or the person is actually costing the company. Because if you assign resources, they spend a lot of time, and this does not get fixed. Perhaps they do not have the skill set. Maybe you need someone a little bit more specialized. I have been in many projects where I would fly in and the customer would basically pay all lodging and transportation to make sure that the job is done right and correctly, because the resources that they have on staff are not sufficient to resolve some of these more complex engineering problems.

What I have also found, and this is not always the case, but in certain instances, if you are in the shoes of an integrator, consultant, or contractor, it is easier to let the end user fail and realize that they need you than to try and convince them before that happens. I have been in many projects or instances where I would make the argument that something needs to be changed or upgraded for one reason or another. It could be obsolescence. It could be the fact that the part is not functioning as expected. It could be to reduce complexity of a specific system. They would, of course, shrug off some of these suggestions. And then a couple of years later, I would get a call. Somebody else was trying to fix that problem. They were simply unable to, and then they bring you in, I would say as a last measure, to solve that problem.

So again, when the customer says he loves to fix it on his own, that is great. Allow them to try that approach. And when things fail, they will be sure to call you. If you try to pressure them into signing a contract, my experience tells me that it is going to be very difficult to convince them of going through you as opposed to if their mind is set on going on their own.

Hi, my name is Vladimir Romanov. I am the founder of Joltek as well as SolisPLC. With a background in electrical engineering and an MBA, and over a decade of experience leading projects in manufacturing and industrial automation, I help engineers, managers, and manufacturers make smarter technical and business decisions, modernize their operations, and build stronger careers. If you are serious about manufacturing, automation, and staying ahead in the industry, subscribe and join the community.

We have a great learning question as the next one, and it comes from the username Jigkal. And he says, "Teaching myself PLC basics." We see the image of a CLICK PLC. And he says, "I am an electrician that is required to screw around with PLCs, and I hate it. I have a few laying around and have been trying to get a basic system up and running so I can get more comfortable with the software and the function of the ports. I have played with the digital versions, but they are not as good for learning field or work time. Does anyone have advice on affordable device kits like photo sensors, lasers, and related equipment I could order for a reasonable price to get my device working and interesting? In the end, I am mostly going to have lights, motors, and VFDs that do not accomplish anything, just set up and function, experience that I cannot get in the field on customer equipment. As for software, is there a PLC that can be flashed to use a variety of software, or would I need to buy specific PLCs to familiarize myself with each type of unit? I know it is a basic question, but even a bit of advice would help. Thank you."

So this is a very good question. As many of you may know, I run a company and website called SolisPLC. I have been teaching PLCs for many years, I would say almost a decade at this point, and this question is extremely relevant to what I have been doing. So I am going to just go back and address some of his questions line by line. "I am an electrician that is required to screw around with PLCs, and I hate it." That is an unfortunate statement. I really hope that he does not truly mean it. I think there is a lot of fun to the world of PLCs. That being said, I have worked with electricians that only prefer to stay with hardware, with wiring. They can even solve some of the mechanical problems. I would honestly, if you dislike this portion of your job, either push through and figure out what might be interesting, or talk with your managers and explain to them that you might want somebody else to help you on some of these engagements, because to outright say that you hate PLCs is a bit of a strong statement.

"I have a few laying around." So he has already got a kit, I am assuming, and he is looking to get more comfortable with the software. "I have played with digital versions, but they are not as good for learning field work for me. Does anyone have advice on affordable device kits like photo sensors, lasers?" So my general advice is that you can purchase very inexpensive hardware as he describes on AliExpress, Alibaba, and eBay. A lot of the industry specific hardware gets sold for relatively cheap on those websites. However, what I will say is there is not a whole lot of value, at least in my opinion. Once you have wired up a couple of sensors, once you have understood maybe a 4 to 20 milliamp loop, maybe a 0 to 10 type of a sensor, you have pretty much done all of it. I do not think there is a ton of information and experience you will get once you have wired in a couple of those field devices to and from the PLC. So I honestly would not spend a lot of money. I personally have a bunch of sensors laying around, and sometimes I will experiment with protocols and maybe some advanced things, but they do not necessarily have that much value once they have been set up.

"In the end, I am mostly going to have lights and motors and VFDs." The value of VFDs, if you want to spend the capital, I would recommend purchasing VFDs that you are going to set up for your customers, so this could be on the Rockwell side, it is the PowerFlex series. On the Siemens side, it is the G120 series. A lot of different vendors are going to have VFDs. You can find ABB VFDs, Danfoss VFDs, and so on and so forth. So there is some value in setting those up. I do not think that I would pay sticker price, as some of them could be a couple of thousands of dollars. Is it really worth it for you to get that hands-on experience? I do not think so.

He asks for kits. And I wanted to give some advice not only on the kits when it comes to buttons and field devices, but on the PLC side. So he has got a CLICK PLC. This is known to be one of the least expensive options in our industry. What I will say is, depending on the nature of your work, so if he is going out and has to deal with PLCs, I would try to, number one, ask your employer to borrow or to basically purchase a trainer PLC that is of the same variety as the one you are setting up. Of course, if that option is impossible, you can either spend your own money on a CLICK PLC. I often recommend a MicroLogix 1100 series PLC that you can purchase on eBay. It is no longer sold as it was discontinued a couple of years ago, but it still utilizes RSLogix 500 free edition. That is a fantastic way to learn.

He says, "Software, is there a PLC that can be flashed to use a variety of software, or would I need to buy specific PLCs to familiarize myself with each type of unit?" And the unfortunate reality of our industry is that a lot of the PLCs are going to be hardware specific. So if you look behind me, I have got the Rockwell, I have got the Siemens, I have got the AutomationDirect Productivity, I have got an Opto 22 groov RIO, and all of them are going to be programmed differently. Now, it is not to say that it is impossible what you are describing, as we are seeing more and more vendors release what is called a virtual PLC. So in theory, you could buy an industrial computer, and you could run a full PLC on top of that computer. It is not a simulation. So just to be completely clear, it is a full blown PLC running typically within a dockerized container that runs the full software from which you can control equipment. It is still in the early stages, so I guess the short answer is I would advise purchasing the PLC that you would normally see or utilize in the field so that it is as close as possible to what you would like to learn on that front. "Basic question, a bit of advice would help." Again, do not hesitate to reach out if you are in a similar situation. Happy to give you some pointers. I know that some of these investments could be quite costly depending on the situation you are in.

We have got a next interesting comment which comes from the username VictorRovorRZ. And they say, "Why is every PLC automation software a buggy, bloated mess? In my experience, I have yet to encounter a software that does not take an inexplainable amount of time to load or have a random exception error with no explanation or outright just close randomly or stop working after a Windows update. Or updating the software introduces new bugs, or the installation fails for some mysterious reason, or you have to install six different software just to be able to use the brand. Looking at you, Schneider. I have experience with Schneider Electric, Mitsubishi, and Siemens, and all of them suffer from similar issues."

And I certainly have my personal opinion on this topic. I left a comment in the section down below. My general thought is that, number one, the software that we use in our industry from all the vendors that he has described is going to be challenging to build for a variety of reasons. Number one, it is fairly low volume. This is not an application that you run on your phone that is going to sell millions of copies where you have a ton of data from different users. Number two, the software is an engineering tool that is quite extensive. You need to start thinking about not just the software component, but also the fact that it interfaces with different ports. There are going to be different tools, from flashing firmware, which is going to be an embedded level system, to building the software and compiling that down to binary for either a microcontroller or a microprocessor, a variety of different controllers at the same time.

I know that at least on the Rockwell side, some controllers will run the version of a microcontroller, some are microprocessors, some are going to have redundancy. So we are talking about very different architectures and thus a lot of complexity. You can talk with those tools over USB, you can talk over Ethernet, RS232, RS485, and Data Highway Plus. All of those drivers are going to be contained within that software. And through a parallel, when we look at consumer grade software, if you were to, for example, release an iOS or an Android application today, in two years, that same application would be completely obsolete. Whereas in the industrial space, most of the vendors will relatively, I would say, support the controllers and components that come from the previous century even. So it is important to keep that in mind as we deploy some of these tools.

To the comment on the Windows side, I would prefer personally that there was more software that is compatible with Linux. The reality of our industry is that for whatever reason, Windows is the OS of choice. However, it is important to understand that Windows also went through a variety of different migrations. So to expect the vendor that has released the software in 2002 to be compatible with the Windows version of today, but also compatible with controllers from the 1990s, is a fairly tall order.

So all that being said, I think that there is a unique challenge for our industry. Can things be done better? I would hope so, but it is of course a challenge for the vendors, for the OS, for the hardware manufacturers. There is a lot to consider when it comes to PLC and automation software.

The next question that we have comes from the visualization side. And so the user by the name MeepMop coming through says, "How to create HMI visuals like the photo," with a 3D generated, what I would expect to be a Factory IO type of an image with some settings, and we can see some kind of a CNC machine representation with a robot that I am assuming is tending for that specific equipment. We have a few conveyor belts, and we have a lot of, of course, safety guards with what I would assume are lasers right at the entrance and the exit for safety purposes.

And I have a couple of comments on this front. I do not particularly care for the nice 3D animations or visuals on HMI panels. Now, the other extreme, so if you are looking at some standards, there is ISA 101, which is basically the specifications as to how control rooms and HMIs should be designed, which typically leans towards the minimalistic side. And there is an argument to be made as to whether or not one is better than the other.

Personally, I fall somewhere in the middle. I prefer more of a minimalistic design, but not as far as the ISA 101 standards go. And what I generally mean by that is, number one, I would want a 2D representation and more, I would say, flat icon visuals for the specific equipment. You can draw out the image right from the front, and you can show different sensors and different areas. You can even show the position of this specific robot. But too many visuals and too many 3D things happening could be extremely distracting. On the flip side, if everything is completely two dimensional, you just see numbers, you are just seeing colors representing the conveyors running, I also believe that the HMI loses some vibrance. So, I would say again, if you were to show up to that specific equipment and just look at the HMI, you should be able to understand what is going on.

So to answer the question of the user, how do you create HMI visuals like the photo? I am assuming that this was a screenshot, like I said, that was taken from Factory IO. You can create all sorts of things in Blender, for example. You can create things in SolidWorks. It is extremely time consuming. So are there companies that have built assets, and can you download examples of very specific machines? Absolutely. Is it worth your time? Is it going to make the HMI better? I do not believe so, unless there are maybe very specific reasons as to why you are going that route, or maybe it is a customer requirement. I do not think that this adds value for the time spent to create these assets.

What I would spend more time on is, number one, functionality of the machine. Number two, thinking about the UI and the UX as it relates to either ISA 101 or, if it is done outside of our industry. I have spent several years and have done projects on Android as well as iOS development. And I have referenced this in the past because those folks have employed scientists and individuals that spent many years researching what makes a good screen, what makes a good UI, what makes good buttons, what color scheme to choose. And I think that actually adds value into our industry because it will make the HMI screens a lot more accessible for the operators, reduce the learning curve, and thus truly add value to the specific system.

We have got a last question, which is interview related. This is important for our industry. I think there are not enough questions being asked like this, so it is my pleasure to answer this user, which is Tiathiste. And they say, "Job requires PLC knowledge. How can I impress during an interview? I am interviewing for a position tomorrow. My friend who works there got me this interview by passing my resume after he heard I was finishing my degree in data science. He said that they need a programmer. I asked the position name and he said data engineer or something of the sorts, but I never saw a job listing anywhere. I asked what I should need to know, and he told me PLCs. I have worked with PLCs a little in the past and am currently brushing up using the resources that are pinned. But if PLCs are brought up, how can I impress during an interview? What would be good questions to ask about their internal systems with PLCs?"

And then he made an edit. I do not know if this was at the beginning of the post or a couple of comments later, and he says, "Okay, to be defined. I did not word it well. I am not trying to lie. I am trying to show that I am not completely clueless when it comes to PLCs. I do not have the experience with PLCs, and they should have been able to tell that by looking at my resume. I do not think they want someone who is creating the PLCs. I assume they already have a lot of people who do that. But like others said, I was able to show interest in their systems and the willingness to learn. I think they want someone who can export and process the data of the PLCs."

There is a lot of confusing information, I will be honest, as I read through that question. This edit actually seems to be detracting from the main question that I had assumed was being asked. But let us go through this line by line. "How can I impress during an interview?" So once you are selected for an interview, and this is a misconception that I used to have in the past, you need to understand whom it is that you are talking with, who is interviewing you. In most cases, you are not being assessed at a technical level. Once you are brought on site, it means that they have reviewed your resume. They understand very well what your experience and knowledge level is. They have already, and you have mentioned this a little bit later in the comment, they know that you do not know PLCs very well. They understand that you have just graduated and at best maybe a PLC was mentioned. I would probably assume that you would take a little bit of time to understand the general landscape of PLCs, but I do not think that they would expect you to come in fully knowing about PLCs and being able to answer questions.

So personally, if I was interviewing you, I would be a lot less concerned about very specific technical knowledge, which we both know you do not have in this specific instance. What I would be looking for is, number one, a person that is very curious, that is excited to be working for this specific team or department. Again, obviously there is no mention of this, but someone who is curious about the general technology of industrial automation, is generally curious about the company, and shows some curiosity during the interview.

We will get into the questions in just a moment. Number two, you want to be very safety conscious oriented, in my opinion. These systems are very different than what you would normally be trained for in software or data science or even electrical engineering. So when I am talking to a newly minted engineer or technician, I want to make sure that they have at least some appreciation of the dangers of going onto the plant floor, for the process, for the electrical panels, and that they can be trusted not to put themselves in danger and the rest of the team in danger.

The next point, very important. I would obviously spend some time knowing the general information around PLCs, but try to find maybe some information as to the company's utilization of PLCs and control systems. I would ask questions such as, what kind of control systems are you utilizing? If you are unable to find that online, I would talk to them about the different hardware solutions when it comes to PLCs, maybe HMIs. I would ask them questions, how do you collect the data from the PLCs right now? Where do you store the data? There are going to be a lot of nuances. I have already made tons of videos on this channel. But basically, you can spend a lot of time just being curious, since you are going to come in on the data science side, about what kind of data are they collecting, what kind of data are they storing, where are they going to store it, in what kind of format, what kind of a historian are they going to use? What is going to be done with this data? Who is going to ultimately benefit from having this information? Maybe they are looking to just monitor a couple of pieces of their own hardware if they are a machine builder. Maybe they want to give the client access to a full telemetry suite of different products. But conversely, maybe they are building out a data capability that allows them to leverage AI. It is all very unclear. So if your bread and butter, and if you are looking to spend more time on data, you should probably ask a lot of questions on that side. I would not try to impress them by reciting some obscure knowledge of PLCs. I do not think that anyone likes a know it all in a very broad sense of that word. You should be asking questions that truly interest you. Talk about different projects. Ask them about what they have lined up in the next six months, in the next year, maybe a couple of years. How could you most benefit the company and organization? What is the learning curve like if you were to join that position? How can you best support them, but also how can they best train you on some of these best practices? All of those questions demonstrate interest and ability, but also the human side of yourself. You are not there to just solve technical problems. You are probably going to work with other counterparts, mechanical engineers, process engineers, other data scientists, managers, operators, and technicians if this is an end user, or assembly workers if this is a machine builder. So you need to demonstrate not only the technical expertise, but at least the ability to communicate with your peers and potential colleagues in the company.