Evolution of Robotics Webinar
![Evolution of Robotics Webinar Evolution of Robotics Webinar](/pic/evolution_of_robotics_webinar/cFVmazduMFdOZFE_.jpeg)
Thanks, everyone for joining today's, webcast, we're. Gonna give. It one minute, for. Any of the scrag lawyers to join and then we'll get started. Let's. Go. Here. To see one minute I. Think. Anybody in the chat, if if, you can hear us or if there's any issues with the connection feel free to you. Know send us a quick note there and we can try, to get that resolved but hopefully everyone can see. You in here as well. This. Will our. Guys I think it will be our first clip first. Webinar without a PowerPoint. This should be good good, conversation. Okay. Well let's, get started thank. You everyone for joining today's, webinar the, evolution, of robotics technology, I'm, your host Jason. Likens I'm the VP of Product Management here at ripcord, and today. Joining, me on our panel of experts. For. Discussion, today are Kevan Hall CTO. And founder of ripcord as well. As saman Farid, who is partner. I do ventures. Kevin. You want to give a little bit more background in introduction, to yourself. Sure. Yeah, thanks everyone, for joining as. Jason, mentioned my name is Kevin, so. My, background is in the robotics. Space and manufacturing. Specifically. Using computer, vision systems. To. Do all sorts of different applications. That, we've worked on in the year at ripcord really. Gearing that technology, towards, the. Digitization process. So. Really excited to be. A part of this and with. Some on here as well looking, forward to talking about a, variety, of topics and interesting, areas for for. This technology both add, record, and also in, other spaces too so, saman. You want to give. A quick intro. Yeah. Absolutely great, too great, to be here it's really really wonderful to spend, time with both, of you guys and I. Hear. Something question from the rest I. Happy. To give a little bit of background. My. My name is Salma and I have been investing, in kind of AI machine learning and robotics for last. 10, years or so through. A variety, of different venture, funds most. Recently I started I headed, up the, to venture fund for Baidu which is. Kind. Of similar to Google is a large search, engine company and also has a large AI, and research initiative, and.
Basically. You know I've been really, passionate about this. Kind of wave. Of Technology. Change that's happening and. One. Of the reasons is that you know we, just see so many different opportunities. For, robotics. And artificial. Intelligence to change the way that everybody works and lives across. Every industry and. Its really we're. Just at the very beginning, of this of this big transformation, but it's going to be like. Another industrial revolution if, not bigger and. So yeah. It's amazing, to be with record, an, incredible, example of this, kind of change. Great. Thank. You and welcome, Kevin and saman, so. Today's topic, that. The you, know it's a very fast pace it's a leading-edge market. Using. All the latest technologies. With robotics, machine, learning AI natural. Language processing all. These technologies, are catching. Up to. Market. Initiatives. Digital. Transformation, and otherwise to, get, some business benefit, and, leverage, that across the organization. Kevin. Maybe you can start off and as, folks, in this new technology of, particular, interest, to ripcord. Yeah. Sure. So you. Know at ripcord we, when we were first getting started we. Started looking at all all. The paper-based, processes. That. Businesses, after, almost every business still deals. With as well as the kind of historical, backlog. Of all of this paper. Data and, as, we view it stored. In warehouses. The. You know state of the art at that time was still. Just using hand tools and, high-speed. Scanners to try to create images of. This. Content and then use software to do. Things like OCR - you know extract, some some text first search, and. Where. We really became interested in specifically. Robotics, and AI from the beginning where it was when we asked. Ourselves is, that as good as it gets or, you, know is, there a better way to do this and using. Today's technology you. Know that we're really able to, you. Know take a crack at solving this problem so, from. From the start that's what we knew we wanted to do leveraging. Robots and different, AI technologies. Both. On the machines but also within the rest of our software, pipeline, to, pull, pull. The most valuable data, rapidly. Out of out of that content so you, know ripcord that's been a part of our you, know DNA from the very beginning because. We you know took this new approach to solving, solving. For that problem. The. Proverbial, building. A better mousetrap more. More efficient. Well. Thanks. And, Sam, on your. End you have a particular. Benefit. Being. In a venture capital, in this type of technology. So. You get to see all the leading edge stuff you know kind of first to market what, are some of those things being out there today. Kind. Of. Active. In the market today and starting, to get a little bit bleeding edge in next generation. Yeah. I think you. Know a lot of the, recent. Changes. That have happened were. A result of - and. A major shifts, that. Happened over the last ten years on the one side I think, the smart phone kind, of revolution, and the mobile penetration. Created. This supply chain where, people have been able to make components. For. Much much cheaper than they previously could, so whether, that sensors, or cameras. Screens. Or batteries, all, of those are things that. Are extremely, useful, when you want to build a robot it. Brings down the cost it, increases. The capability, and. So, that's been you, know one one of these kind of trending, forces that we've seen on. The other side. Around. Kind of deep, learning as a technology, specifically, with broader into AI the. Ability for computers. Can interpret, complex. Sets, of data has also drastically increased in the last few years and. That's resulted in things like natural language processing computer. Vision. OCR. You, know all of these drastically. Improving, and we see can themself chiming cars and many other things happening. As a result of those two major. Forces and so, you. Know to answer your question about what we're seeing specifically.
A. Very. Large rain. Where. These technologies can become useful. Whether. It's kind of document, management it's, and and processing. Of paper records. Or. You. Know automating. Your yourself. Driving car automatic. Automatic, entire. Manufacturing, lying, or facility, you, know creating. Much more efficiency in a construction site. Surveying. You know using. Drones to solve all kinds of problems I think kind of the list is is endless and to be honest I think we were only seeing just the beginning of it so far I think as, these. First few applications. Take, off. We'll. See a new generation of components, and a new generation of technology and a lot more interest being poured into this space and. So really it is kind of this exponential, shift. In in. The way that, basically. You know tasks, are accomplished, that decisions are made in. Every. Kind of business. Yeah. The. Pace is amazing, it seemed like we're on this cusp of the. Technology. Going. To push us through to, the next levels. Of things that we never even thought possible before. Its. Fourth generation of, the Industrial, Revolution it's, it's a very exciting time some. Interesting. Things to just consider in terms of like human, robot interaction and. What that looks like you know we're I think today I'm gonna talk, you, try to keep it focused on like the enterprise right and and how wrote this technology is great for the enterprise and you. Know when you think about businesses, it's going to be like a lot of those concepts, or commas you. Know what does that interaction, look like especially as we see more and more of these use cases come along. Let. Me yeah pretty pretty exciting. To watch. Yeah. So it's so Kevin to expand on that a bit. There's. This concept, of you, know even even the title of this webinar. Robotics. But I think, people use that term very broadly. There's. Robotic. Process automation, as an example the. That. Is, more a software, box yeah and. Then, there's the physical hardware of, an, actual robot. How, do you see these two technologies come together in, the, enterprise and with, kind of this digital transformation. Unstructured. Content what. Are some of the use cases and, the combinations, of those two technologies. Yeah. I, think both, are, both. Forms of kind of core. Hardware, robotics, and machines. And automation as well as you. Know software boss through, through, RPA are. Really. Going to be critical for for. The enterprises I think our PA in particular, is. Super. Interesting because it can globally. Support most. Most. Companies out there will have a use case or do right now have a use case for that, sort of technology to take the you, know where robotics, typically.
Really Shine is in more the more environments. Where the really repetitive tasks, are very. Tasks. That have like a bunch. Of data where you need some just a robotic. System or an automated system behind. The scenes crunching. The numbers and, building. Those models so I think you. Know that's the case for so many different enterprises, out there today so I think what. We're really excited about is we you're. In rapport think that our. Our. Approach, is really enabling for, our PA because, there's. A huge amount of data, out there that companies, have stored. On paper, and. For, machine learning to, really. Become. Take. Take off and, for these models to really improve. To help classify your records and extract valuable data specific. To your organization, it's. Going to be really important to connect. That information. Through a pipeline, that's, leveraging. Machine, learning to you improve, and. That's what we've you, know designed, our system to do from the beginning which is only going to better enable. Software. Automation, then, do. More admit you know, more. Advanced processes, as well with with the content. Know. That. Yeah. So the automation, piece so. Using. Data-driven automation. Extracting, some some values, out of the content, and and, driving. Some automation, processes, a, lot. Of people are that, RPAs expose that very well people. Are taking advantage of it do you see any areas where maybe. The technology, could be used in other ways. Maybe. Some underrated, or underutilized. Capabilities. Of. Its. Enrichment unstructured. Content. Kevin. You, guys you. Know what's. The next wave of taking. Advantage of this or what. Would have people not quite realized, yet. Around. Analytics, maybe or line of business integrations. Things like that. Yeah. I mean I think it the easiest. Way for me to think about it is when I when I try to consider you know in enterprise sort of the types of decisions people need to make of. Course every every, team every department, with an organization, has different kinds of decisions that they're trying to make. And. You. Know so for example with, the accounting team you know they. Need to know about all, the accounts receivable payables. All of the inventory. On hand like, that was if you're managing a warehouse and logistics in, your in your facility, you, need to know about every. Piece of equipment. That's in there and, periodically. Need to be able to do surveys and to be able to predict, demand whether. It's on the marketing side whether it's on a manufacturing. Floor and and across you know every department, in and. You, know a large organization, decisions are being made. The challenge, is that you know a lot of the times I think people, don't, have enough information to make the. Most optimal, decision they, follow a certain set of rules for example. They'll. Order inventory. When it falls below you, know a certain threshold, those, types of decisions right now. Service. Well when. We were doing things in. A very, very repeated, repeatable. Wait so if I was trying to build you know 10 million chairs you know then you know whenever I get below a certain order quantity I'll order, a few more you know more steel or whatever supplies, that I need but.
As We move to a world where complexity. Is increasing and for example a manufacturing, facility or, a, company. Company, really needs to adapt. To their plans on a daily basis, rather than making, annual plans and then just sticking to the annual plan or a quarterly plans instant the quarterly plan really, organizations, need to basically be updating their plans on a daily basis spaced on lots. And lots of information in. Order to be able to do that the, amount of information that people will need is just so much more and. The. The reality is the majority of that information that they need is not currently. In a digital form so, whether it's on paper or whether it's just in the form of the number of boxes that are sitting on shelves in your warehouse, there. Needs to be some way of taking, all, of that physical, world and turning, it into a digital representation so, that we can understand, it, understand. Much much more of it and then, be able to make decisions using, it. That's. Thanks. Summon. Kevin. Maybe you can apply a similar, use case. To. Leveraging, that paper the, information, locked away in paper in the oil and gas industry, I know some of our attendees, are from, that space and and, maybe. You can share some of your experiences, with that. Yeah. Sure, so we, I, mean, right now probably behind this wall there's a bunch of robots scanning. Some some paper for the oil and gas industry. You, know it's. Been pretty pretty. Fascinating, to see. See. The sorts of applications and, insights, that. These. Companies, are looking to get out of their historical, data, and. What's. Just fascinating. Is you. Know if you can get over that in that, really good to the realization, that you, know the data you have on. These records when. You get. Through. A pipeline that's going to be able to learn. And improve as, it sees more and more of that content, whether, you're trying to extract information about. Location. Like where, was this what well, does this record tie. To or. You. Know deeper, insights into actually, on exploration, side trying to look at the combination. Of inputs from, active data, today as well as historical, data that's been in paper to then go and do. Varying. Levels of exploration. Looking for new new places. To drill, for oil, it's. We've. Seen a really strong, commitment, from our. Customers in this space to leverage. Both you, know new, new information, that's coming in through, from. Out on the edge and from their different IOT devices and other data tracking systems and pair, it with their historical, data which. There's just a huge amount of useful, information there and. If you put all those things through the same training. Pipeline that, what comes out the other side is really. Powerful and only gets better with time so, you know the early adopters, I think in this case are gonna be the most in. The best position moving forward I think. That's seeing that I just want to pick up on something that Kevin mentioned because I think that it's a really really meaningful. Point that you know people are, implementing. You know let's say IOT devices or, sensors all these different things on all of their equipment but, a lot of the. Historical. Structure like you know what. There, takes, a lot of time to manually, create. A structure for all that data to go somewhere, it, becomes very inefficient, and people often, don't have a good like it's very very difficult to justify the ROI of those, of, those investments in in new assets, but. Once, they have a certain baseline of for example taking like a lot of historical information and having a digital representation of that and using that to build these, models then, things, like IOT, data and sensor data can make a lot more sense because, it can tie in to, this structure. Of. You, know all. The history of the, facility, and I think that applies just, as much in the oil and gas and maybe. One more addition. On that on, the a little. More on the hardware robotic side it's been really exciting, from. From our point of view you, know one of the things you you often, get with a robotic, machine as opposed to you know more traditional like automated. Machine is is flexibility. You. Know robot very inherently more flexible, than a you know a more rigid machine. Or mechanism, and this. Really. Speaks. To I think some of the things. That are only recently becoming. Possible. In, the space of you know using robotics and AI, to, handle, more. Powerfully more. Types of inputs, and, what, we've found when we started processing, records for the oil and gas space is that there's, all sorts of different formats, that. That. Data is stored on you. Know everything, from regular sheets, of paper it's a really long well. Log, big. Maps and that sort of thing, with. With different fasteners and all, of the above like a huge, amount of variable. Input and using. A robot with. A perception, system on the front end that's able to you.
Know Robustly. Figure out what we're looking at which is a you, know for us ml, based system. We. Were able to like rapidly. Address. Those sorts of new types of content, which. Is something I think a few a few years or five years ago five years ago when, it wouldn't necessarily been, possible because it has the leverage technology, that need new technology, that we have today both. On the perception, side but also then. On the hardware side of the machine that's flexible enough to handle, that type of content and. We see that it's. Really exciting because this, type of application I hope other companies. Will see. A similar trend where. They're able to you, know use robots, to unlock. And. Disrupt a certain space which. Then opens up a lot of other opportunities to build more machines and more robots to keep, keep. Improving, and, we see a lot of opportunity for that in the oil and gas space. Yeah. It, seems like the deeper your data set is, the more. You. Know information. You have to create your machine, learning, or AI and. You. Know if 80%, of, the world is unstructured, and a large percent of that isn't paper. Getting. Access to that information to feed your datasets and and just. Make those things smarter, each time is, is really. Valuable and important, you. Mentioned something kind of an interesting. Kevin, about the. Ability of the robots to handle the paper and, can. You explain a little bit some you know what the benefits of that is related, to privacy. And security and you, know some of these GDP are, type. Regulations, that are top, of mind. Sure. Yeah. I think, you know part of one of the biggest benefits of using a machine to automate. The digitization process. Is. The. Fact that you know it's it's much less touch time from, people which has a multitude. Of benefits, including. Including, speed and, efficiency that, we can we can process and the scale in which we can do it and. From a security point of view you. Know there's less less. Opportunity, for or. Less people looking. At the records there's not necessarily, a KPI or a metric in the industry that tracks that met that, number but for. Sure using machines to to. Automate the process from, a security point of view is is really important, and. I think that's a trend probably, for a lot of different, uses. For robots in the enterprise like something. That comes to mind is like in a hospital, if you have ground vehicles, delivering. You. Know important, like medicines, or the you know patient, information from room to room which, is a common growing, use case in healthcare, you. Know that's pretty sensitive information and, there's. A benefit to using a machine to do that sort of that. Sort of workload from, an efficiency point of view but also from a privacy point of view because, you. Know it's it's a machine to robot it it is. Just gonna go and do want to do what it's told and not. Not. Have the opportunity for that security, issue and. I think just to add to that the other other benefit, of it is that there's you, know very very clear, records for everything that happens, so. If. Something. Is processed in a certain way or came through a certain box or whatever it might be exactly. What touchpoints, happened and when they happen where they happen it's all reported, in. A lot of detail which, from, a privacy perspective and, the security perspective is extremely, extremely valuable. When. I want to go back to one of the comments you made about, you. Know how some of this mining. The unstructured content. Can. Add value to other use. Cases in, the case of IOT it makes that much stronger so. If you look at kind of data out there today you have, structured, data in databases. Coming. From applications, you have unstructured. Content. From. Users. Emails. Documents. These types of things and, you have device data. That. Is very. Quickly escalating. It's kind of you know as far as the next generation this is where a lot of the data is going to be coming from IOT. Can. You could you maybe come up with a use case on how some. Of the structured, unstructured content. Could. Could, enrich, and add value. For that IOT. Data, that's coming in how. Would an IOT, you. Know something's, coming up for maintenance. On. My. Brakes on my car and. Right. Leverage some other unstructured. Or structured content to combine with that to. Make that IOT data that much more valuable. That. Was an interesting point you made, yeah. I think you, know like the easiest way to think about it is the historical data tells, you what the. Kind, of normal conditions, should, be or what the optimal. Conditions should be so if, you just hook up a sensor to whether it's your car or.
You. Know some some pump in a refinery. It's, very very hard to know whether the. Data that you're collecting means, that it's it's an it's in good shape or it's an or, it's, having issues or it's trending, towards some, issue and. The only way to be able to and. Contextualise that data is to, have historical, records so that historical, records i of course some. Of it may be records, with that same data some of it may be records. Of other operating conditions that give you correlations, for example when, yeah if, they I can be trained to understand, that, a. Certain, sound in the car is correlated, with certain. Types of. Engine. Failure well, that by itself has a very very useful piece, of information so, even if it doesn't necessarily, detect the. Engine failure through, some sensor that you install if you have a microphone in there and you detect a sound, that is similar to the sound that happened when there was previously image failure that pattern, is a useful piece of information for, the operator so. I think that you. Know that that, contextual. Side of things really comes from structured. Data or some, human annotation, right there has to be a person, who comes in and says you know this is what we're trying to optimize we, want it to, make this type of sound or something like that um and, and the only way to get there is is to have. Sufficient. At historical, records and you. Know I think like machine. Maintenance is. One example but. Extent. Data. That we need for example when. Trying to make decisions about you know where to where to where. To drill for. A new new. Oil well or what. Type of land to buy a land, use rights to buy or not buy a. Lot, of that type of historical, information right. Now there's people who look through it as geologists, and things like that who, have a, certain amount of you know subject matter knowledge and certain lot of experience, but, that, often could be supplemented, by lots. And lots and lots of training examples and you know even the best engineer. In the world you know has maybe you know 60 or 70 years of experience, that they can bring to the table. You. Know but but AI system, trained on disparate. Data from a large number of sources can, easily you know in a matter of days train on hundreds, of years of experience across, hundreds of locations and. So that by, itself gives that gives a company that kind of structural advantage of experienced. Revenge decision makings it's a huge use. Driver for why, to get your content digital right because, you need, it to be digital then give your teams and software tools. The opportunity, to even you know train, or learn from that that you know important in history I think you something, else like maybe a slightly different angle to think at this from or, something that we really think about a lot of report. Is, this concept, of you know a fleet a fleet of machines or robots and. Being. Able to you, know work together and learn from one another based on their. You. Know the history of what they've processed and, what they've seen this. Is especially. Important, I think for any. Robotic. System that's having to deal with, an. Unstructured, world. Or input, you. Know at ripcord this is you know mainly in the form of unstructured, content where, we have sheets, of all different sizes and, fasteners. And folders and all sorts of things that have that, we've used to store information, and needing. To be able to flexibly handle that but if you know and all sorts, of different use, cases whether it's you know robots that are trying to pick. Up something or know where to put something off of a table, or out of a bin or into, a truck you, know there's there's. Some really interest, there almost. It's required for those systems to work together and, learn, from one another, you, know based on what they're seen in the world and, that all is coming out of this broader data set that, you. Know these systems are you. Know connecting, to. So, that's really important in a record we we see you know we. Have this fleet of systems and we're building, more and more as. We grow I mean, it's really.
Interesting To connect these. These. Systems together and share that data to to, then make better decisions. And. Get, even more automated, and more flexible. Yeah. I think. That's an interesting point, and you. Know maybe, we're, part. Of the next generation goes, you have this catalog, of information, that you can cross-reference and, find. Other. Relevances. And relationships. And. You. Know scores. And ranks of, the content, that. Can be leveraged in ways, that aren't really thought about today but you have all the the data and you can use it in the future, kind. Of like Salman, was mentioning earlier it's a little bit it's. Much more relevant to get an IOT piece of data if you can compare it against ten years worth of previous data so. That's very interesting so. Next. Question maybe final question you, know with all these new technologies, and, leading-edge. Stuff and you. Know moving faster, than ever before what's. What, some of your predictions, for the future where do you see this all going and how, it's being leveraged, you. Know other than robots taking over the world and in. That would, he what do you see saman is uh in. The future what, do you see in your crystal ball I. Think. There's. Really just so many different. So. Many different areas where these kinds of changes are happening I think. Right. Now it's been very very exciting because we've seen you. Know internally we think about it as node efficiency. Versus system efficiency, so right, now a lot of nodes in. Any. Business are becoming more and more efficient because of Technology so whether that's, optimizing. You know ads on your online platform. Or whether it's, increasing. The efficiency of a, certain. Process in your manufacturing line. You. Know these each of these nodes there's that there's equipment there's technology, and there's ideas that are coming together to make that more efficient. What. We think is really really going to be exciting is in, the next few years as enough, nodes become digitized, and have. Some technology built into it it's, going to become possible to have system level efficiency so. It's just a mobile efficiency, and you know what I mean by that is when. You want to link different parts of a very complex chain to each other. So you know you manage your your supply chain and your procurement in combination, with your. Expected. Kind of production. And you run your marketing efforts and all, of these things work, together. So. That you, can kind of optimize, so when you know for example you're going to have some spare capacity a, month, before that your marketing efforts can wrap or. If you know that you, have you're. Producing something and there's some or, some. Short, supply on then, in, advance your procurement system will automatically go out and start meeting across, a variety of replacement. Materials, so. I think that you know like the entire enterprise is, getting to a point where a. Lot. Of these managerial, functions, are going to be able to be drastically, enhanced, with. Technology, and. Humans, are going to be able to really unleash the, creative, side of that so. Instead of spending time just kind, of doing very repetitive work, or a rule-based work, we'll. Be able to to. Spend our time kind of creatively making decisions, and thinking about new ways, that we can build our businesses. That's. Great that's exciting that it, kind of gets on the edge of the predict and prescribe, and, you're. Not our mate. Thank. You for that yeah. What's. The future hold. Yeah. Good and, so on I think that's an amazing. Amazing. Future and that's, I think really aligns. With you, know a lot of my, thoughts and what I see so, maybe all a. Little, more on maybe the robotics, or the manufacturing, side.
I. Think, we're going to be in manufacturing. A pretty. Major. Leap. Forward, through. Connecting, our. Many systems. Together and having that more you, know global, perspective. Of the system as a whole and leveraging. You. Know artificial intelligence, to become more flexible which. I think will be really really, important. I also. Think you know I, think, we see a lot and some on you probably I think you're invested, in many, of these types of technologies to you know enabling, technologies, for this space I think we'll see a lot of growth in you, know this is things like sensors, and, communications. Platforms. And things like this that are able to really. Enable, the. Future of. These. Robots to then get into the, enterprise for instance and. Start. To provide this more universal, value. You. Know for all sorts of different use cases so I think you know we'll see a lot of interesting. Technology, coming out of that area as well. That. Things like. Lyde our new cameras, and, you. Know things that are you. Know edge compute, and all. Sorts of different areas and really. Be enabling technologies, for this space and then you know companies are gonna bundle. Those pieces together to you know build really impactful, systems, to you. Know support. Certain. Industries and and also more globally, the the. Enterprise as a whole. Yeah. Just just because you mentioned cameras, you know one example just to kind of take that point home is really. The. Way that all, these devices are going to be rebuilt for for, having computers, to, interpret, the physical world so right, now cameras, are pretty much all designed in a way that are optimized, to make you know the pixels look pretty and have colors and look, really nice so that when a human eye looks at that that music and interpret it understand what that means but. You know as as more and more of this data is being fed to computers, cameras. Are going to be able to be completely redesigned, so. That cameras, are built you know for example with multispectral. Lenses. That look at a wide, array of different things that look at depth they look at it like you mentioned lidar and.
That By itself is going to create just, a whole new revolution, of what, robotics is able to do. It's. An exciting, time and, I. Think we're all looking forward to be a part of that in the future. So. I want to thank our guests saman. And kevin for. Joining, and sharing their thoughts on, the. Evolution, of robotics, technology. This. Webcast will be recorded, and rebroadcast. Thank. You everyone for joining and, have a great day. All. Right thank you. Yeah, thanks, I'm on I appreciate, the conversation. Great. - yeah, really enjoyed it. All. Right thanks piers. Yeah.
2019-02-17 14:54