Privacy Policy
more

Taxal – Design for the IoT

Taxal – Design for the IoT

[00:00:07] Hello, my name’s Alan Behrens, and we’re here to speak about design for the Iot, when we talk about the Iot, most of us speak about things such as Output’s big data, how to store information, how to analyse it, AI machine learning and how to store it, of course. But really, you know, how do we know that’s the right data we want to gather? You know, do we know if it’s the right information for the right type of product? Do we know it’s reliable? Do we know that the data collection is safe? So our show today, we’ve got two distinguished guests. We’ve got Chris Dickey from Technologies. Hello, Chris. And we’ve got Marc Sampson from Siemens. Hello, Mark. And what we’re going to do is just have a chat about this particular topic, which is actually a very prudent question to ask. So what is what is the rationale between the right information for the right types of products and is it reliable and safe? So let me just hand over to Chris in the first instance, just for him to explain who he is, what his company does, and a bit of background. Thank you, Chris.

[00:01:25] Thanks, Allan. So my name is Chris Netsky, I’m the CEO of Petchem Technology. We develop a software product called the Maintenance, the Way Design Environment. What it does is combine a number of the engineering analyses that are required for design and sustainment of complex products with a focus on how you can put condition based maintenance into those into those products. We’ve been operating for over 10 years. Most of our customers are aerospace and defence and another patient and safety critical systems. And in the last couple of years, we’ve started with Siemens globally.

[00:02:04] Great. Thank you very much, Chris. Mark, how about yourself? Just a little bit of background.

[00:02:08] Well, good morning. This is Mark Sampson. I’m a system engineering evangelist, as you would call it, and work on integrating NBFC with the product lifecycle, including products liability, essentially. Our goal is here is to think about it in a preventative way to design our products for safety, reliability, rather than trying to react to the problems of recalls later. So I guess I’m looking forward to this conversation. I’m located outside of Las Vegas, so I apologise. I’m not quite awake yet.

[00:02:36] So not a problem. Well, it’s obviously because you’ve been out too late, so we won’t hold that against you. All right. So so the first question I’ve got to ask is what distinguishes products for use in the Iot? What types of products are we talking about here, Chris?

[00:02:57] So I think the first thing I put is with the IoT, it’s what questions you really trying to answer? I think this is a big difference between people that are looking to solve problems around maintenance and reliability and safety for complex systems. You know, they need to have a much stronger design focus to understand, you know, the physics of failure, the actual engineering principles that underlie the behaviour of those systems and what they need to do to be able to monitor and interpret what that what that behaviour is going to result in five years down time maintenance and so on. So I think it’s important to talk when we we look at the the IoT, we’re focussing very much on the engineering aspects of that complex safety critical systems and being able to make sure that you can design and maintain them more effectively, which is different, I think, from, you know, people that are looking to do consumer behaviour applications from Internet traffic.

[00:04:02] So we’re talking about aeroplanes, ships, missiles, very complex, sophisticated systems.

[00:04:12] Yeah, you know, I look as well, cars, trains, power stations, anything that this complex engineering systems involve and also that a safety and mission critical, because I think that, you know, that’s one of the things with the Iot that people a lot of what you can learn from operational data. But it’s important to remember when you’re talking about these types of systems, you kind of have a failure.

[00:04:39] You have a failure in a nuclear power plant. That’s a pretty bad day. You have the seven.

[00:04:45] Know you want to make sure that when you’re designing IoT applications for those missions, safety critical systems, you really need to understand the engineering underneath to inform what those Iot applications might look like.

[00:05:00] OK, so so so essentially what you’re saying is you designed for four complex systems. Variety is going to be different to just designed for other things, including designed variety, because you’re talking about the safety criticality and the inability to suffer any degree of failure. Is that right, Chris?

[00:05:22] Yeah, look, I think, you know, the common example people have is that their fridge is now hooked up and it can automatically order milk when it’s down and so on and so forth. Now, you know, if you have a failure in that type of system and you don’t get the milk in the morning, you know, again, that might be a bad day for you.

[00:05:37] But it’s not the same sort of bad day that you have if you’re your Kaizo or automatic braking kicks on in the middle of the freeway and, you know, you come to a screeching halt in front of oncoming traffic. So I think, yeah, it’s important to understand when they might mission and safety critical systems really need to engineer that IoT application.

[00:06:01] I’m just sorry.

[00:06:03] I didn’t mean to interrupt. Mark, you’ve got some examples, haven’t you? Have.

[00:06:07] Well, it’s not just it’s just not those kinds of systems you think about that are safety critical. It’s the little things that often will get you right. So, for example, you guys remember, I guess it was about a year or so ago where we had the TVs, the Internet connected circuit, closed circuit TVs that were generating a denial of service attack because someone was able to hack them by default. Or, for example, the latest one I see here now is the fact that the Nest thermostats are turning themselves off and have to be connected to the Internet. And this was last winter time in the middle of the winter time. So it was shutting down heaters and stuff like that. So it’s just not it’s just not the complex systems that you’re used to. It’s also these complex little things that you’re now putting in everywhere. And they have interesting interactions that that end up coupling, if you will, and causing other kinds of problems along this process, like the cameras with the denial of service attack that they were producing.

[00:07:06] Right. So so so the challenges that we’re talking about don’t just affect individuals, they affect businesses, because also there’s this degree of business risk failure, especially in the aerospace and automotive industry. You know, if you have a catastrophic failure, we’ve seen this in the past, it can be obviously reputation damaging, as are many of the car recalls and car problems that people have had over the past years. So there are some huge business ramifications on the.

[00:07:40] Yeah, so, for example, you’re familiar with a cell phone manufacturer had a mistake in their designs or that ended up costing them think the number was I get the full number for you, but the one that comes to mind here in the United States, for example, last year there was 37 million recalls. That’s down. This is on automotive automotive vehicles. That’s down from 55 million the prior year. The National Highway Traffic Safety Administration here in the US that keeps track of this estimates that it’s about a hundred dollars per recall per vehicle. So you can imagine that’s five and a half billion dollars that was spent in the automotive industry. Plus, they estimate that this was an interesting statistic, that 12 people died taking cars in to get fixed from these recalls. Oh, so anyway, it’s it’s expensive anyway, you look at it.

[00:08:32] Yeah. Yeah. So so let’s talk about some of the design. What are the differences in designing for the IoT as opposed to just designing Chris? I mean, what’s the differentiation that.

[00:08:51] Well, I think picking up on Mark’s point before the complexity of these systems is very much due to the fact that there’s a lot of software, so cyber physical systems that will mechatronics, if you will. So it’s important that you can understand what are the things, the physics of fire that will lead to the greatest state of operations for those systems where, you know, potentially things can go wrong. So it’s important to understand those things as you go through the traditional design process. But now there’s technology that exists that allows you to use that same information to decide where you might put sensors. So you go through a process, how can this system fail? How can you know, what are the consequences of that failure? You can assess those particular risks and then you can say, well, if these are the specific technical risks that I need to be monitoring, what are the sensors that I need to be able to monitor? And so this is I think typically it’s known as a fusion approach where you should be designing from a physics of failure. The underlying engineering principles need to be understood through your design process, and you need to involve the people that will be doing any sort of diagnostics or maintenance to make sure that you can put in place a system that will give those people the right information as they need it so that they can make better decisions about when to do things and when to schedule it.

[00:10:18] Right. So how does one get started with doing? I mean, we understand the problem is very sophisticated. One needs to plan and strategize to actually accomplish better objectives. How do you get started?

[00:10:34] Let me.

[00:10:36] I was just going to say I had something for you, Alan, that what Chris is describing is kind of a different perspective or a different way of thinking about this problem. When I was a new engineer, because I was the youngest in the group and the least senior I got I got strapped with delivering bad news about one of these kinds of situations in the product we were in process of developing. And I remember after the swearing was over, that would have been in English our prior conversation. But when this hearing was over, the the the programme manager looked at me, said, Sampson, why don’t you give me something? I can see this problem coming right now. It’s like driving a car, looking through the rear-view mirror. I can see the problem after I run over it. Yeah. So I think here is to provide something to do with what if the types of technologies that you could see the problem coming and line your product either to reduce that risk or do something directly about it so you can avoid the problem altogether.

[00:11:29] Right. So, Chris, what about the special skills and training and attributes that a company might need to acquire to get going in this? Tell us some of the things that one needs to consider to to actually apply these types of technologies.

[00:11:46] So I think certainly it’s the engineers that are involved wherever you is, you have to engineer an IoT application.

[00:11:55] So you certainly need the engineers of all the various types. And I think it’s also important that you want to be able to take the data off of that. So, I mean, they’re calling them data scientists now. But people that have maths or the big data capabilities to be able to be able to interpret some of that data is coming through. But, you know, again, for the IoT applications that we’re looking at, the same thing, mission critical systems, we think the engineers have a lot to say. They’re ones that are going to be determining what you need to monitor and why. And.

[00:12:30] OK, great. And also sort of the help that companies might get from people like yourselves. I mean, are there any other organisations that they can go to for advice on the top on this particular domain or topics?

[00:12:47] Look, I think this should I mean, not as good as us, of course, for Australia, but there are obviously a lot of people that a lot of money being spent, money always attracts people. So there are a lot of people that are doing this. I guess one thing that’s interesting is that you are really looking at a continuation of engineering processes and methods and techniques that have been around for 30 or 40 years, and the technology is enabling that.

[00:13:15] So I think it’s prudent to take a look at what people have done in the past and apply those methods with these new technologies. And there are rather a lot of I.T. aspects to this. So certainly the I.T. community has a lot to add, but we think engineers are pretty important, too. Again, the questions we want to answer.

[00:13:38] OK, and what about the Arawa? Because obviously everybody these need these sort of new methodologies and the tools, obviously, that you’re providing and others providing within the domain cost money people need to justify them. How does one go about justifying the ROIC or the expenditure?

[00:13:56] Well, that’s one of the problems we’ve got generally on is, is the fact that the accountants are busy looking at the rear-view mirror for money you used to spend that you’re not going to have to spend in the future, and that’s how they can engage things. What we’re after here in this particular case is adjusting our thinking to to think about stuff. We only have to prevent one or two problems to pay for ourselves pretty rapidly. So, for example, was the one, the washing machine that exploded from the modal vibrations, you guys can probably hook that one up if you apply the hundred dollar per recall kind of of factor to that. That’s let me see, something like 2.8 million washers that need to be fixed. So that’s two hundred eighty million dollars that we’re talking about saving. If these people or the designers had placed a sensor or done something about it before it got into people’s homes. So there are ways about preventing future problems which are very, very costly, billions of dollars in some situations. And some of them kill companies. Right. Like the like the airbags and others that are out there. Right.

[00:15:01] Right. Chris, you got any interesting anecdotes and on return.

[00:15:08] Yeah, look, I think, you know, picking up on that point, certainly that’s the big argument is it’s all about the arabized, typically based on cost avoidance, which is something that sometimes the accountants struggle with. But but again, I think if you bring this approach that, you know, you’re not just going to sprinkle sensors all over a system and then see what data comes back and see how you can use it if you’re talking about. Commercial aircraft, for example, any piece of equipment that goes down has to be certified, and so there has to be a business case and an engineering case for any sensor that you put on those sorts of systems. So a really I guess what we’re saying is you need to do the right analysis, the engineering analysis. Combine that with an assessment for whether any particular sensor can buy its way onto the platform. What’s the data that it’s going to give you and what problems will it potentially solve? Because, you know, you have to understand what you’re trying to achieve with the Iot. I think that’s a very important point, too. If you’re if you’re Arawa is based on cost avoidance. It’s cost avoidance of catastrophic failures and optimising in maintenance. And, you know, with a lot of the systems we’re talking about, that’s 70 to 80 percent of total ownership cost. You can shave a couple of percentage points off that overall budget. Then you are talking seven, eight figures potentially depending on the system.

[00:16:29] So, yeah, it’s definitely, definitely, definitely very big numbers and clear case for investment. So so you sort of on a final point, you know, in summary, just give us know why why should people look at contacting yourselves? Because obviously the results of this conversation is if people are interested to obviously contact yourself, Chris or Mark, why why should people and and after that, obviously somebody on on the concerns that you’re trying to address.

[00:17:08] Well, you know, certainly I think Michael talked more about saying this, but I think the interesting thing about stamens is they provide an end to end solution. They’ve got all the bits and pieces, all the various technologies of which we are obviously one part. Why do people want to talk to technology? Why use made? Because it allows you to engineer IoT applications. It allows you to to do what we just said before, understand what those particular problems are and highly complex, interdependent systems. Be able to assess and identify those risks and be able to put in place a solution that leverages what the Iot offers, cheap data, being able to communicate it safely and interpret it and so on, but giving focus to what that data is and focussing really on asking the right questions so that the Iot can give you those answers. So, yeah, that’s I think what I can bring to AIX.

[00:18:04] Mark, any comment from yourself to what we’re after here is a Stephens’s in the end solution, as Chris pointed out. And we’re bringing an important aspect with Crucis technology into the product lifecycle for people to see in a forward way potential issues that should be addressed before it gets too late. We end up with a recall costing us lots of time and money and reputation. So I think that’s what’s important here.

[00:18:29] Great. Excellent. Very interesting discussion. Thank you very much, Chris and Mark. I think those who are watching will agree some interesting and valid points designed for the eyes. He really is different to just design. So making sure that we have the right information gathered and that safe and reliable is ultimately the objective, especially with complex and sophisticated products and and with human interactions or safety is concerned. So thank you very much. And I hope that you enjoyed the session. Thank you.