<img alt="" src="https://secure.leadforensics.com/211043.png" style="display:none;">
  • Listen to all episodes on
  • Spotify
  • iTunes
What is WITS?

We’ve made it to the last episode of Season 2. Ken and David discuss the components and significance of WITS.

This episode is sponsored by

Gibson Reports
Gibson Reports

Featured Episode

Episode Transcription

Hi, I'm Ken Miller.

And I'm David Erdos.

And welcome to the Erdos Miller New Technology Podcast. We talk about all things drilling tech. Dave, it's the last episode of the season. So Season 3 will be coming shortly. Big thank you to our sponsor for this season, Gibson Reports. So check them out at gibsonreports.com. Excellent NWD and directional drilling market share reports. Been a huge, huge help for us. And thank you, David. Anyways, so we're just going to get into it. Today's topic is WITS and I'm just shocked that we haven't talked about this yet. I was going through the list of episodes and I was like, "How could we not have talked about WITS?"

Yeah.

Like, "Did we just forget it?" Right. So.

I guess so.

And before we get into it, a big announcement: We're hiring.

Yeah.

So go check out our website for all the roles we have available. We're always looking for really talented and passionate people to help us change industries. So wits.

Yes.

So like I mentioned, I just couldn't believe we went through the whole list of episodes and hadn't done an episode about WITS, right? So what is WITS? Or do you know the official term?

I have it pulled up because I didn't know.

Read it to me.

Well site information, transfer specification.

Also known as the protocol that just won't die. Right.

And also just works.

It also just works because it is a testament to simplicity.

Yeah.

Right?

Mm-hmm (affirmative).

So what is this weird WITS thing. Right? So this was one of the first projects I had in the industry. Right. And WITS is just this really simple, really old protocol for having one machine talk to another machine at the rig site, right.

Point-to-point.

Point-to-point. One machine to another machine and that's it.

Okay.

And it's still used in every system out there. Right.

Yeah.

Every new system. Well, we got to have a WITS port because we have to talk to other machines. Right? And it's so old. They talk about the oil field being slow to adopt new technologies. It's still based on an RS232-

232.

Serial port, right? It's still like a 1980's serial port, right?

Yeah.

It's like ... Even today, you go out and you build this brand new control system for oil and gas, NWD drilling, whatever. And you still got a plan to have an old school RS232 port in there to talk WITS.

Yeah. Virtually no computers have that anymore. You have to have a USB serial adapter. Some older industrial computers do, of course.

The last generation like laptops we have from Dell did it. But like these new ones they're trying to copy-

Only on the docs.

Yeah. They're trying to copy Apple and get all cool and less ports and that port has got to die, right? It's super old. So yeah, it's used commonly to exchange data between machines. I'll tell you one thing we do every time. We go out and we hook up an NWD surface system to whatever the EDR system is. Right.

Right.

Electronic data recorder. As far as I know. I've heard other definitions. Right.

I think that's correct.

But this is the machine and the set of computers and sensors that monitors the rig on surface. Right. So this is commonly provided by Pason and TACO or ... I'm going to get in trouble for the guys that I'm forgetting. There's a few other ones out there. But those are the big ones that I interface to a lot.

And so those measure top drives, speed, pumps.

Everything, right?

And everything.

Right. And so when we go out there with an NWD system, we need to get the depth from the rig and maybe the RPMs and all this other kind of stuff, which we're not going to take the time to rig up our own sensors. Right. And so the pace on with the TACO system, the EDR already has that sensor information. So we're just going to connect to it via a WITS port and that's it. Right? And then we'll start exchanging data, right?

Yeah. So does the EDR have a bunch of WITS ports that everyone ... You have like five or 10 WITS ports and you can just plug in as many systems as you want?

They charge for them.

Oh.

You can get more, but they do charge for them. Right.

Yeah.

So it's not like a USB hub where it's super easy to expand, but there are multiple ports. Right.

Yeah.

And so, but yeah, and then that's how it works. Now, where WITS starts to break down is when we have more than two machines needing to communicate, and it gets a little complicated.

Sharing the same data stream.

Well, they have to pass the data through or whatever else. Or you have to just get more ports and I'll use the EDR system as a hub or whatever else. Right.

Yeah.

But it's a very simple protocol. Like I said, the physical layer is the RS232 port. So it's a DB9 connector, right?

Yeah.

It looks like your old school, serial port for your printer or whatever. Right. And it's a very simple protocol. You can actually go to our website and get, there's a basic tutorial about the protocol. And there's some information on the specifications. We can put those links at the bottom in the video. And you know what? It's really kind of a very good, it's a wonderful project when you're first getting in the industry to write a WITS protocol.

It teaches you about the parameters that are of interest in the industry and it teaches you how to do some programming stuff. It's something you should be able to figure out in about a week. Right. And get that going. Now, I do have a story about that. So I was a young software developer in the industry, and doing a WITS interface was one of my first projects. Right.

Mm-hmm (affirmative).

And I found this old website that had all the WITS documentation. Right. And it's pretty simple protocol, but it was also kind of strict in the way that it was written. Like floating point numbers should have this many before the decimal point and this many after. And that's it.

Because it's asky.

Huh?

It's asky.

It's asky, right? It's all textual based. Right.

Yeah.

So what's the Unicode again? I'm kidding. But, and so I wrote all this code. I spent two weeks writing it and I went out to test it with my first machine. Right. And I was so proud of it because I had adhered completely to the specification and I expected my code to work the first time. Because I played by the rules. It should work. Right. I go up there and I try to connect to a Pason system and it doesn't work at all. Like, "What's going on here?" Right. And what I came to find out is that nobody plays by the rules. They all took the specification as a guideline, just kind of did it a little bit flexibly, right?

Yeah.

This many data points before the ... This many decimals before the floating point. This many ... It changes from value to value, whatever. Right. So you get a little more flexible. We all agree that there's data packets and what the IDs are in the data packets. But then what comes in the payload is a little bit=.

Kind of squishy.

Nebulous. It's squishy. Right. So I had to take all of that hard checking code out of my stuff and make it a bit more flexible. And then it worked.

Yeah. It's interesting with all these, well, oil field protocols, they seem to be sometimes like a game of telephone between the different manufacturers. Everybody hears a little different thing and says things a little differently.

Yeah.

With the protocols.

Well, it's funny, a lot of the standard ... I think WITS is one of the few standards that's actually published that we all adhere to. Right.

Yeah.

Because everything in NWD that ends up being standard is just standard because we copied it off of one guy back in the day.

Yeah.

Everybody kept it the same. It's not like we all agreed on a specification, right?

Yeah.

So, okay. If it's so old, why hasn't someone replaced it, right.

Yeah, that's a good question. Probably because of simplicity.

It's dead simple. It's really simple, right. WITSML is a different thing, right?

Yeah.

And what WITSML does really well is it gets data off of the rig site back to a server. Right. And so that a geologist remotely can look at it and see what's going on and you can aggregate all the data. It's really good for getting, collecting the data at the rig site and then putting it in an archive and then exchanging that between different companies or whatever else. Right.

Mm-hmm (affirmative).

But what's interesting is the first version of WITSML didn't really address the real time component.

Hmm.

And so, because they didn't really address inner machine transfer. Or at least the way that they did address it, wasn't popular.

Yeah.

And so I know that WITSML 2.0 has a new thing that they're trying to aim to replace WITS. I haven't really looked into it. But there are people that are trying to come up with a new protocol. Right. Because we have a lot more sheens on the rig site today. They need to exchange a lot more information and they need to know more about the information. Right.

Yeah.

Because the problem with WITS is it makes a lot of assumptions.

Yeah.

It assumes that whatever you send it happened right then and there.

Right.

And that's it. Right. And there's no other way to really tell it anything else, like, "Oh, this happened a second ago." Okay. No, no. There are no messages. There's no [inaudible 00:07:49] or whatever else.

No guarantee of that.

So I think we have to ... I think that looking at the WITS protocol and looking at how simple it is, is something that every engineer should look at and really strive and look back to is like a great example of KISS: Keep It Simple Stupid.

And so that's really great. But we do need a new protocol. Right. And so we've been doing some experimentation. We have our NetWITS, which is basically WITS over a network layer so that we can talk over Wi-Fi networks and have multiple things talking to each other, but we kept the simplicity of WITS. I'm sure there's other people out there trying to do stuff.

Yeah.

And so that's it. It's definitely a long podcast. WITS is a simple protocol. Right. So for machine to machine data transfer to exchange information between these real time machines at the website. Right. So what other questions should we talk about?

I think ... I have been wondering why WITSML and its successor hasn't had greater adoption yet?

Well, I think popularity and simplicity. That's my opinion. Right?

Yeah.

But I can definitely see a benefit to having something networked where you can have every system talk to every other system and not have to have some fixed intermediary. And it's kind of limiting in one way.

The rigs have gotten pretty good about having a Wi-Fi network locally there on the rig, right. Now, the other thing that we're going to have to address as an industry is a lot of people are trying to build things that want to control the rig. WITS only handles status.

Data exchange.

Data exchange. Like what's currently happening? There's no control mechanisms. Right?

Yeah.

And there's ... If you say, okay, I built this machine. I think this machine can drill. I think it can control the top drive and whatever else. You can drill any better than another automatic driller or better than a human drill or whatever else. Right.

Yeah.

How do you interface that to the rig?

Right. It's got to be custom for every implementation.

Well, I can tell you right now that the first thing that's happened is that whoever owns the rig, their legal department is going to say, "Absolutely not." And so I almost wonder if one of the things the industry doesn't need to collaborate on is some sort of central hub black box that allows multiple machines to talk to it and governs the data.

Yeah.

And so anything that has been built and tested with this mythical black box is guaranteed to be safe. And the black box itself does a little bit of safety checks so that you don't tell the top drive to spin at 10,000 RPM or something like that. Right. But just kind of this trusted third-party exchange ... Maybe a physical box is not the right embodiment.

Yeah.

But something like that, to where every manufacturer going out and testing against every other manufacturer is saying it's really-

Not practical.

Not practical. Right. It's very tiresome, it's complicated, it's expensive. It's time consuming. If we all just collaborate on this one central point of focus, and took into consideration data exchange and control and all these other things. And we all agree to just build and test to that interface, then maybe we could have things like a small team, which is really important to me. Enabling small teams who are really innovative to do big things. Right.

Yeah.

Because right now it's just difficult for them to get past the legal aspect, even though they're really brilliant technologists. And so I think could be a really interesting invention that might level the playing field.

Yeah. I definitely think, one thing I've noticed is people are generally more collaborative than you might think.

Yeah.

Even competitors. Just like with ICWSA. All these different companies, that are, technically, competitors, coming together and working-

[crosstalk 00:11:16] For the good of the industry.

For the good of the industry and developing standards and recommendations for how to improve.

Yeah. There's no reason that couldn't happen on the data exchange side, too.

Yeah.

Yeah. Okay. Well, I'm out of stuff to talk about on WITS.

Unfortunately, I am, too.

Yeah. So we'll stop there for today. So remember this is the end of our second season. Season 3 is coming up next. And thank you again to our sponsor Gibson Reports. Check them out at gibsonreports.com for excellent NWD and directional drilling market share reports. And we'll see you next season.