Episode 115: Transcript
Seeing is Believing: Manufacturing Processes Under the Lens
Speaker 0 00:00:04 Make it right. The manufacturing podcast,
Speaker 1 00:00:09 They say that seeing is believing. And when it comes to finding out where something’s going wrong in your manufacturing process, it’s important to be able to identify quickly and easily where things went off course recently, a us company landed $25 million in series B financing to push their factory video analytics to the next level. The company’s currently working with a variety of companies in the automotive medical device and electronics industries. Welcome to the makeup right podcast. I’m Janet Eastman. And this week on the show, we look at the value of video on the factory floor with my guest, Dave Prager he’s head of marketing and business development at drishti. Welcome to the Make It Right podcast. Dave, thank you for having me. It’s my pleasure. So, uh, I gave an overview there, but, um, you work day to day and this stuff. So explain just how drishti works.
Speaker 2 00:01:02 I think in order to understand how drishti works, it’s important to talk about what’s going on on the factory floor without touristy and everyone in the audience has probably got an industrial engineering background and is probably stood on the factory floor with their stopwatch, trying to figure out what’s going on or trying to figure out what happened. And that’s really the way it’s been for about a hundred years, you know, a hundred years ago, Frank and Lillian Gilbreth, they really pioneered this concept of time and motion studies and, you know, Henry Ford and, and, and, and, uh, all, all of that activity a hundred years ago in making the first mass manufacturing was all predicated on this idea that you can take any work and you can build a set of standardized work, uh, with time associated with it. And then that gives you the foundation for finding improvements.
Speaker 2 00:01:51 So a hundred years later, factories in a lot of ways are very different than Henry Ford time. Uh, but in a lot of ways, they’re still exactly the same. We’re still a bunch of people running out with running around with stopwatches, trying to figure out either what’s happening or what has happened. So that’s where we got started. Uh, our CEO is a guy named Prasad Akella 30 years ago, Prasad was working at GM. Uh, and then he led a team that developed the first collaborative robots. The task was, we’ve got people putting in these, um, these really heavy doors is really heavy dashboards and they have to contort their bodies and lift and carry. And it’s really difficult. Robots would be great for this, but robots lack the precision to guide everything it’s really at the time, uh, to guide everything to where it needs to go.
Speaker 2 00:02:38 So the solution then was to combine man and machine combined the best of technology, the best of people, and the result was cobots collaborative robots, which is a huge industry today. It first started from his team. So for 30 years, Prasad spent his time thinking about how do I combine the best of a person and the best of technology and create something than its whole. So four years ago, five years ago, I first started thinking about this problem in the age of AI, he said, Oh, well, you know, when I was in the factory, I spent all my time out there with a stopwatch. And I can’t believe that problem in the year 2016 is still not solved. Uh, so that was where just you got started is how can we take technology to augment what the process engineers or the quality engineers the supervisors are doing to make them more efficient?
Speaker 2 00:03:31 Uh, so the, the short answer to your question is that Trisha is a video analytics and video traceability solution. We put a camera and every workstation, we send the video stream to the cloud and we give the people on the floor access to two things, three things. Actually, we give them access to the video itself. So if I want to know what happened Tuesday at 9:00 AM on station three online for, I just put it in and it’s as easy as using YouTube. And suddenly I’m seeing exactly what happened. So root cause analysis is not, it’s not hours, it’s not days it’s it’s three minutes. Um, and the second thing we do is we give you tools to add data on top of that stored video. So if I want to re run a time and motion study, I can just run it, pull up a clip, a video, put in my cycle, define it, put in my steps to find those.
Speaker 2 00:04:18 And suddenly I’ve got my time in motion study done, and it’s in a fraction of the time that you would spend normally doing it. And the third thing we do is we actually have an AI that in certain circumstances, usually high volume lines, we can actually do portions of the time study automatically. So we can actually, in some of our customers, we’re recognizing the beginning and ending of every single cycle. So you can imagine how much time engineers are spending, typically hunting for imbalances, hunting for variability. They know there’s a problem. They know there’s NBA. This don’t know where now, if they had every single cycle with defined beginning and end and they had videos, you could see exactly what happened in those outliers. Suddenly solving all these variability problems becomes very, very simple. So that’s where we start video plus analytics plus data.
Speaker 1 00:05:07 Wow. Okay. So what is the penetration of video on manufacturing lines right now?
Speaker 2 00:05:13 Well, you know, um, for years, I really remember this four years ago. Prisaad so I’d worked with Prisaad previously. Uh, so four years ago he called me up and said, Hey, Dave, we’ve got this really interesting idea. Um, come meet me at a Starbucks. Cause we met at the Starbucks and he told me about his idea for drishti. And the first thing I said was, well, surely every factory has video cameras. This is, this is a waste of time because everybody must already have this right. Put up a GoPro and you’re done. And the first thing he said to me, it was no, no, you’d be shocked. You’d be surprised. And I was very disbelieving at the time. And now I believe it there’s very, very little video inside a factory. And the reason why is it’s not just as simple as putting up a GoPro because you have to maintain the, you have to make sure that the data goes somewhere and you have to get that data out of the system.
Speaker 2 00:06:01 If I’ve got a thousand little thumb drives with video from station exit, you know, Tuesday at station X from last Monday and station X from three months ago, it’s going to be some amount of time to get to the video and find the second one I’m looking for it. It’s a, it’s a prohibitive amount of time, even if it’s five minutes per thing, it’s still too prohibitive. And so people just don’t go down that route because you just, you’re looking for a needle in a haystack. You need this whole video infrastructure built around the video camera in order to make it usable. Or another way to think about it is you have to take the technology and integrate it into manufacturing workflows. If you’re not giving the data in a way that the people know how to use and are comfortable using the, you might as well not have it at all.
Speaker 1 00:06:48 How long did it take to develop this?
Speaker 2 00:06:51 Uh, it took, um, probably a year or two of working with one or two customer. Actually we can reveal them. Now we’re working with flex tronics and Denso, uh, to really understand how the technology worked even for ourselves and the use cases that were supported and, and what’s really going to create value for the customers. Um, so those early days we had a, a vision for using AI to recognize what a person’s doing in order to create data and use that data to improve the line. It was a, it was a very broad vision, but it was very kind of ill-defined in a lot of ways. And over the course of the early lifetime of the company, we learned more and more about the, the extraordinary amount of use cases. It could be supported when you have continuous data on human activities and when you have video to back up every data point. Um, so it took some time, but, but now, uh, we’re definitely in market with a tool that is creating significant value. And it’s, it’s very exciting to be part of something where it grows from a crazy idea that you talked about at a Starbucks to a, you know, now we’re a company that just closed our series B and we’ve got 50, some people, uh, working very hard towards this shared vision.
Speaker 1 00:08:05 So let’s talk about some of the key challenges that your clients are addressing using this technology. Can you share some stories?
Speaker 2 00:08:11 Yeah. You know, it really depends on the kind of production you’re running. Um, and what we’ve learned is that the use cases in low volume lines are very different than the use cases in high volume lines. So let’s talk about low volume line for a second in a low volume line, uh, your primary problems going to be NVA, finding out where your waste is. Uh, it’s going to be quality because low volume usually means high and cycle times high cycle times usually mean lots of operations. And the more operations you have, the more mistakes are likely to get made. Uh, and the third problem you’re going to have is in training because you’ve got operators coming in and out, and especially in the time of COVID, you’ve got more absenteeism than normal. Uh, you are going to struggle to teach new people to do the process and you throw in a high mix and every unit has 10,000 variants.
Speaker 2 00:08:58 And then you’re really going to struggle with helping people understand what to do. So just having the video alone has been very transformative in these different kinds of, in this environment. So to address the MVA, uh, here’s a great example. Um, one of our customers is running about 10 minutes cycles, uh, three stations in their line, 10 minutes cycles, except for they’re actually running about 12 or 13 minute cycles. Because when you look at them all together, there’s a ton of hidden waste and they cannot, or they could not figure out where the waste was coming from. Because of course, every time the process engineer would go to the line, the line would be hit perfectly. And so what it took was taking the cycles that we recorded and actually going through them at high speed and just starting to Mark roughly where the waste was.
Speaker 2 00:09:45 Okay. At three minutes and 20 seconds, the person left the station and at three minutes and 60 seconds, a person came back. So they do a rough pass of going through the video. They started categorizing the waste, and then they can go with the finer tooth comb. Well, here’s 17 instances on Tuesday of a person leaving the station. Why? Well here, you know, 13 of the 17, they were carrying tools, three of the 17, they were carrying parts and one of the 17, they were carrying nothing. Okay. It looks like we have a problem with missing tools. And just being able to go through a large dataset where it would take a person sitting there for hours and hours to figure this out normally, and be able to very quickly go through all these cycles and figure out your Prieto of problems. You can solve that really, really quickly.
Speaker 2 00:10:28 You would never, they would never know that a tool shortage, uh, until they had this evidence. And they’d never get this evidence because it never happens when you’re looking. And same thing with quality. When that defect comes, when the, you know, God forbid the customer calls back a few months later and says, Hey, this part isn’t working well, you can go to drisha, you can type in serial number one, two, three, four, five, and then you’re watching the exact assembly of that part. And you can say, Oh, we forgot the flange. Our fault. We’ll replace that for you. Or we can say, you know what? The flange is there maybe got damaged in shipping. Actually a lot of our electronics customers have problems where their end customers will mess up the unit and then try to blame the factory and they’ll return it always missing the screw or that part was broken when we got it. Well, suddenly instead of processing the RMA and taking the loss, uh, we can actually, you know, our customers can actually use Rishi to look at the state of the unit as it leaves the factory. And obviously there’s a lot that can happen between the factory and the end customer, but they can start to narrow it down. And that kind of information is actually changing. How you think about warranties and returns and even recalls in an age where suddenly you can see the assembly of every single unit.
Speaker 1 00:11:35 Hmm. Okay. So you’ve given me a great picture. Now I get how you can kind of promise an ROI within eight weeks of deployment, because your discovering things about a process that nobody’s going to even think.
Speaker 2 00:11:50 Yeah. Yeah. Thank you for doing the marketing for, I’m glad that, you know, as the, as the marketing guy, having someone say the words I wrote makes me very happy. Um, but it’s true. Uh, so, so what we say is that within eight weeks you see ROI and the ROI starts at day one. So, so let me give you a little bit of the chronology of how that works. Um, day one, you turn on the cameras, let’s say your factory is in battle Creek, Michigan. Well, you’ve got a bunch of product designers who sit in Detroit, uh, who don’t necessarily want to drive the three hours to make sure that their product is being assembled correctly. So they can just log into the video and see it that minute, that second and suddenly you’ve got live view, or, you know, more recently in, in, you know, in this age of COVID, uh, every person who’s on the line is a person who is creating risk in terms of social distancing, the more people who can use a camera to do their work, uh, the safer everyone in and out of the factory is.
Speaker 2 00:12:49 And so we have a lot of customers who are seeing that ROI in terms of the health and safety benefits and, and helping operators feel more comfortable coming back to work. Um, and then beyond that, like I said, uh, anytime there’s a defect or an abnormality, you can see it and right away, there’s no magic. So, so there was an article in the times, um, probably about a year ago now about the golden state warriors back when they were good. I don’t think they’re good anymore, but they were good at one point. And I talked about how at halftime, they would spend half time looking at video of what happened in the first half and making corrections. And they would come out and they would dominate the second house. Um, and that’s exactly the idea. If I can show an operator, an instant replay of a mistake that they made, uh, then they know how to fix it. And more importantly, they’re not going to argue with me and they’re not going to say, Oh, I didn’t make that mistake. Tony, over there made that mistake. It becomes a very unbiased and unemotional way to say, Oh, look, this happened, let’s study the instant replay and correct it for the next time. And then we start to see companies really significantly improve their defect rates and shorten the duration of training just by integrating video into their workflows.
Speaker 1 00:14:03 I think that’s, it’s really interesting because honestly, if you’ve got that camera going and you’re saying, well, you’re doing something wrong here. That person has no idea probably that they’re doing something wrong until maybe they see it on that camera and go, Oh, I see. Because I’m doing this that’s happening. So yeah. Well seeing is believing, right.
Speaker 2 00:14:24 You know, there’s a story that came out of one of our customers recently, where there was a lady on the line who, uh, so some of our customers, we actually embed people there to help our customers succeed. And, and in this particular case, the drifty customer success representative had gotten to know this lady very well. This, this line operator, she was very proud of her work and very meticulous and never made a mistake. And so one day the QC station found a mistake. That was an issue that would come back to her station. And, you know, the supervisor approached her and she swore up and down that she never makes mistakes. And it must’ve been somebody else. Uh, well, they went to the video and they saw that at that particular operation on that particular unit. And by the way, it takes 30 seconds to get there.
Speaker 2 00:15:07 Uh, she reached for the part, she saw that the bin was a little bit askew. She fixed the bin and then neglected to grab the park. So her muscles were working on memory and her muscle memory is you go there and you come back, but she had skipped the part and instantly she says, Oh, Oh my gosh, I didn’t even know. And what would have been a very emotional confrontation and a very upsetting confrontation for her. It became a simple learning moment. And she said, okay, well, that won’t happen again. And, and, and it hasn’t when you take that and you multiply it by a thousand times a day and you start to see real change very quickly.
Speaker 1 00:15:45 So drishti is used by automotive electronics companies, medical device companies. Um, what, I mean, what are they doing in a big automotive factory or are they doing the same thing in pretty much every, every manufacturing process? Really?
Speaker 2 00:16:02 Yeah. So a second ago I talked about the low volume lines. We’re going to focus more on finding their waste, uh, improving training, improving quality, high volume line has it has a different challenge. So, so they’re, um, they’re much more focused on finding the variability and finding where the bottlenecks are and finding why is the line slowing down or what’s going on in the line. Um, and that’s something that we’re actually able to help with in a couple of different ways. And of course you’re focused on efficiency and Kaizen and improving. So let’s talk about, you know, like, like I laid out in the beginning, lean today, the standard way of, of finding significant improvement in align is you have a Kaizen. You have everyone drop everything they’re doing for a week and spend 10 hours, 12 hours a day focused on the line, measuring, observing, gathering data, coming up with ideas.
Speaker 2 00:16:53 Uh, and then after five and 10 days, you have a new process that you then run and, and hopefully you declare success. Um, the first thing that we’re doing is we’re making that process a lot simpler because instead of spending five days, you know, the first three of the five days gathering data, the data’s already there in a high volume line. Our AI can actually do cycle time, every single cycle, which means, you know, before you even show up on the, on the line, not only which station is your problem, or which process is your problem, you also have watched a video of it. You know, you sat there in a classroom and you’ve watched cycle after cycle. Maybe you don’t need this crowd, the operator observing you don’t need to spend a day observing. It’s all there. And then you just use your creativity.
Speaker 2 00:17:35 And this is what engineers love. They don’t have to use their stopwatch, which is the boring part. They get to use their brains and they get to solve the problem. So instead of five people working 10 hours to solve one problem, you can have three people working five, or you can have five people working 10 hours to solve 10 problems. And so there’s this multiplier effect that happens, especially in the high volume lines. I think it’s really appreciated, but that’s so step one is you take the Kaizen and you make it shorter and more efficient. But step two is you realize you don’t even need the Kaizen because if I’m taking continuous data, if I constantly know where are my bottlenecks and where’s my variability, and if I can even slice and dice that data, show me shift one cycle time, performance versus shift two, or even show me rotation one versus rotation to show me Monday versus Friday, you can do all these things with the data because you’ve got the data set that that’s big.
Speaker 2 00:18:28 So suddenly instead of having to do a Kaizen every three months, you can just do continuous improvement, true continuous improvement. You can adjust and try new things. You can say, alright, for rotation one, we’re going to try it this way for rotation tube. We’re going to try it that way. I’m gonna go back to my computer. And by the time lunch is over, I’ll know which way is better. And we’ll get that half a percent. Now, half a percent isn’t much, but if I can do half a percent every couple of days, then again, you start to see real significant improvement real quickly.
Speaker 1 00:18:59 Hmm. So, um, Kevin Snoop, who is the person who sponsors this podcast with, uh, he’s written a book called make it right, five steps to align your manufacturing from the front line to the bottom line. And the industries he really works with are, um, uh, feminine hygiene and diaper products. So these are factories where wow, these things are made super fast and when something goes wrong, your waste can add up really, really fast. And I’m thinking, um, you put a camera on these lines. You’ll when it comes down to raw waste reduction, how quickly can your cameras identify that problem and cut that waste in half. I’m thinking about, you know, what’s the end result after a year you’ve installed this, this system and then you’ve cut your waste by how much
Speaker 2 00:19:46 Yeah. You know, some customers, um, we have seen 20%, uh, potential efficiency improvement, you know, within days, meaning that we identified it. Uh, obviously, you know, it’s, it’s, it’s manufacturing. So you still have to do the work to realize the improvement. What we do is we point you in the right direction. And in fact, you know, a lot of people say, what is AI? Uh, and a lot of people think, well, AI is something that can do anything. Um, really what AI is is, is really, really good, but not perfect pattern recognition. So the way we use AI is we teach AI to recognize patterns that are relevant to people in this case, you know, the, the unit entering versus leaving the station. Uh, and then we use the AI and the other tools to sift through these massive volumes of data to determine what the human being needs to pay attention to.
Speaker 2 00:20:38 So we have one customer where we give, you know, we, we did our cycle time measurement on thousands and thousands of cycles. And then able to say, guys, these are the ones you should pay attention to. And for us to take 10,000 cycles and find 20 that are most impactful to the problems we’re trying to solve, it’s an extraordinary thing. But if you think about the, the diaper lines and the feminine hygiene lines, they’re a little bit different. So drishti is for discreet manufacturing. It’s for assembly lines, station work, primarily bench work today. If we look at process manufacturing, it’s a little bit different of a problem where you’ve got a bunch of people tending machines, uh, and something goes wrong and you have to figure out what activity either just then or in the past led to that outcome. And there’s a huge body of IOT work, where they can measure machines and get data off of machines and figure out there was a problem on the, you know, the, the dispenser at two or 3:00 PM.
Speaker 2 00:21:39 And then all the engineers rushed to the line and, and started doing root cause analysis and trying to figure out, well, what could happen? Was it that guy upstream one minute earlier that guy up shooting 10 minutes earlier? Or was it a person forgot to load this, you know, 20 minutes ago or a person forgot to oil it three weeks ago? And there’s all that investigation is still required. So we have a partner called TBM. And when we first met TBM, they said, they, they, they looked at what we were doing. They were very happy with discrete manufacturing. They said, we want to take you into process manufacturing because they knew that we could be the missing connection between what happened and why it happened. And so if you were to ask Ken, if you were on this podcast, Ken would say, uh, you know, get these guys into a process manufacturing environment and let’s see what they can do. We haven’t done that yet because we’re still a fairly young company, but we know that that’s definitely on our path.
Speaker 1 00:22:30 Okay. So let’s talk a little bit about the 25 million in series B that you recently landed. It’s, uh, from my understanding to pursue AI powered production. So, uh, what does that look like based on the future compared to what you’ve told me so far?
Speaker 2 00:22:46 Yeah. So first of all, 24 of the 25 million, uh, got me a house. So I’m very happy
Speaker 1 00:22:52 And it’s a very nice house.
Speaker 2 00:22:53 Yo. Yeah. Yeah. It’s on stilts and it can walk. Yeah.
Speaker 1 00:22:56 I’ve Googled you and everything. I know exactly what it was.
Speaker 2 00:22:59 Yup. Thank you. Uh don’t don’t tell anybody, well, actually you can’t tell people cause I’ll just tell my house, stand up and walk away from them. No. So the, the, the investment, um, which, you know, we’re obviously very excited, everybody that company’s excited, especially for those early employees like me are, it’s just such a humbling and gratifying moment to have people vote with their wallets and place their faith in you. Uh, it really feels very good. Um, but the purpose is because we are proving value now is the time to extend and scale that value and make sure that more and more people can benefit from it from, so from a business standpoint, it’s to, to get into the market, to get more customers, to prove ourselves and to really, you know, very slowly or hopefully very fast to take over world. Um, but from an R and D standpoint, uh, you know, we are one of the few companies in the world that is parsing streams of video to identify human actions in real time.
Speaker 2 00:23:58 And let me double click about two or three things in that sentence that I think are worth saying, um, people now, if a robot does something, a robot does it almost identical every single time there, even when people are performing repetitive process, they do it differently. Um, and there’s high variability in who people are people of different sizes, different skin colors, left handed, right handed, big, small, different genders. Uh, all of these things make it very difficult for computer vision to traditionally recognize what a person is doing. So for us to have an AI that is able to look through a stream of video and say, okay, there’s a guy starting a cycle, and there’s a lady ending the cycle. And to do it with a very, very low latencies, the data can flow to a supervisor right away, uh, that that’s really cutting edge.
Speaker 2 00:24:47 We don’t talk too much about our technology because we like to focus on the use cases and on the ROI, but, but the technology is under the hood is extraordinary. And that said, it’s still very early days for this. And we have a really visionary roadmap of what we can do with technology that can recognize human actions. And then you start to think as, as the technology is matured and extended, uh, you know, all of the different ideas that start to open up in this environment. Um, it really gets me very excited. It certainly helps me, uh, get up in the morning every morning at 5:00 AM before my kids get up, someone’s gotta help me get up that early that’s for sure.
Speaker 1 00:25:26 Yeah. So Dave, for some key takeaways, um, when you’re looking at, uh, the video solution, what are some key takeaways? You’d like people manufacturing leaders, you know, to think about about this process and how video can work for them.
Speaker 2 00:25:41 Yeah. Good, good point. So, um, the one thing you haven’t asked yet, and we always get this, you know, executive talking about it is how do people feel when they’re put under a camera? Yeah. And you know, if I showed up to work, um, well, first of all, I’d love to go back to the office, but, uh, if I showed up to work and there was a camera, but my desk, and I didn’t know why, and I didn’t have any value to being derived from it, I’d be very upset about it. Um, we don’t have that problem. We actually find that operators become the biggest fans of drifty. And the reason why is that in a data starved environment, the operators themselves also lack data. And that actually hurts them in a couple of different ways. So one, uh, if your job is to follow standardized work every single time, every single cycle, if you do your job, you’re invisible.
Speaker 2 00:26:31 Nobody actually knows you’re doing a good job. And there’s no source of data that says, this person is extraordinary at this job and should become the trainer or should it become the supervisor should become promoted. And it becomes a very frustrating environment where you see the people next to you don’t do as good of a job getting paid the same. They’re getting the same recognition. It can be very upsetting. So the first thing that operators appreciate is suddenly there’s a way for them to prove the value that they’re creating. Um, the second thing that operators really like is that there’s a source of truth. Oftentimes on the line, when there’s a problem, the operators didn’t get blamed and the operators typically know the process better than anybody, but when things go wrong, they don’t always get believed. This was the human emotion, a supervisor or a manager has to very quickly say, you know, this is what’s wrong.
Speaker 2 00:27:20 I’m going to fix it. And if they’re wrong, the operator gets upset. Like we talked about earlier. So having a source of truth to prove that they know their jobs and that the problem is whatever they claim it to be operated for really, really like having that protection. Now, some people say that, you know, in certain environments, people don’t want their work to be exposed. And it is true that people who aren’t trying to do a good job, that that will become apparent, but the people who really are trying and the vast majority of them are, uh, they appreciate having that extra data point, uh, to, to make it. So, and the third thing that they really appreciate, uh, is being able to contribute back to the system. Like I said, the operators know the system better than anybody, and they know where the weak points are, but it’s hard for them to demonstrate back to the engineers or back to the management, what needs to be done to fix the problems suddenly if there’s continuous video and continuous data creation, that becomes very easy for them to say, Hey, I’ve got a new idea.
Speaker 2 00:28:17 Sure. Let’s try it out on rotation one. Oh, that was a great idea. You just saved us three seconds per unit. That’s the kind of thing that we start to see happen from the operator point of view.
Speaker 1 00:28:26 Hm it’s fascinating stuff. And you know, I’m, I know people are concerned about their privacy. I’m concerned about my privacy, but we’re on camera all the time. Right? Like that’s true. Camera’s everywhere. So, um, yeah,
Speaker 2 00:28:40 No, we, don’t sorry. We don’t do facial recognition. We don’t actually know who the person is. Um, when we try, when we can, we try to have camera angles that don’t capture the person because we’re not focused on the person we’re focused on the process. Um, people do ancillary really get seen in the camera. Um, but once they understand, and by the way, that’s probably the most important thing is before the camera goes up, we explained to them what’s going on and we help them understand how they’ll benefit from the camera. And, and ultimately when the company’s investing in software like drishti, which is designed to improve human processes, they’re investing in the person as opposed to automation. And that means that their jobs are safe. And, and, and certainly in this environment, uh, to, you know, today’s economy, uh, people really appreciate the investment being flowed towards their viability, as opposed to, you know, making them obsolete.
Speaker 1 00:29:33 One final question, just how hard is it to understand the data that’s provided? Like, um, I know you’ve, you’ve talked about how simple it is to get to it, but is there a big learning curve when you first implement this to get a sense of, okay, this is what we have to do in order to understand what the video and the information is telling us.
Speaker 2 00:29:53 Yeah. That’s a great point to end on. Um, we’ve very deliberately designed this to take the output of very complex technology and make it very simple and simple and usable. Um, so the video search engine is, is, is as simple as using YouTube and the line balance charts and the, uh, Yama Zumi charts and the, uh, histograms. There are all things that are utterly familiar to the engineers because they use them every day. We’ve been very deliberate to make our tools line up, that their workflows they’re already making decisions based on visualizations of, uh, line downs, for example, that’s great. We give them the data in the form they’re used to, and then with one click, they’re watching the video. And so people very quickly understand how to use it and understand how to integrate your sheet into their workflows.
Speaker 1 00:30:42 Wow. David sounds really cool. And, uh, I wish you guys a great success and I hope we get a chance to speak to you again on make it right. Thanks for joining us. My pleasure. Thank you, Dave Prager is head of marketing and business development at drishti and, uh, you can check them out at, I believe it’s drishti.com. Am I right Dave? Yeah.
Speaker 2 00:31:02 D R I S H T I.
Speaker 1 00:31:05 Okay, perfect. That is our show. This week, please check out our Twitter and LinkedIn feeds that are on our podcast page. You can subscribe and share this podcast with your friends and colleagues through iTunes, Google play, Stitcher, Spotify, and YouTube, and the make it right. Podcast is brought to you by Kevin Snook. He’s a leadership advisor and author of the bestselling book, make it right. Five steps to align your manufacturing business from the frontline to the bottom line. I’m Janet Eastman. I hope you have a great week. Thanks for listening to Make It Right.