Streamline pipeline integrity data management for animal science

How to create outlook signature

good morning good evening uh thanks thanks everybody for for joining in now i'm rahman i'm the global head for iot phi now and uh let me let me extend a very warm welcome to everybody who joined in uh we have uh a lot of people from our group companies and a couple of our customers we are expecting you know a few more to join in but let's let's dive right in right so today we have we have a very special webinar and a very special guest right the webinar we have is is how we are helping organizations use anomaly detection to improve their machine operator and thereby plant productivity right we spoke with a lot of customers across verticals and the feedback that we kept getting was hey everybody wants to do these cool things right i mean industry 4.0 smart manufacturing getting closer to zero on planned down time but but what was lacking was the how part right a lot of organizations leaders asked us hey can you help us answer and answer the whole question we we know what needs to be done we just don't know how to do it and uh so today you know uh prashanth who is our executive advisor uh uh with iotify on industry 4.0 would speak to you know some of that right so let's let's dive right in right uh we'll try and keep it short right and uh and you know open the forum and then for q a right uh the webinar would be available uh as a recording so you know feel free to share feel free to you know now let me take let me take just a couple of minutes to talk about i85 right so we are a part of the simcon group and we have been in industrial automation space for for close to close to three decades now we have been connecting outdoor assets and what that means is that you know today we think we are really good at doing the harder parts of iot web so if you look at the whole iot value chain there are some parts that are really easy to do things that you do on past platforms visualization and all of that the harder parts is something that not many can right and that's that's the area where where we where we play and when i say harder parts you know i mean things like sensors smart platforms and ability to bring machine learning you know from the cloud closer to closer to your high value assets and all of that in terms of verticals you know we we have been very deliberate about vertical speed plane like manufacturing oil and gas utilities water and gas and some work in civic services like smart lighting and water management right so with that you know uh let me let me introduce prashant to you as i already said you know he is um our industry advisor on industry food auto and and smart manufacturing key guides not only so so so a big part of his role is guiding iot5 itself right so he guides our strategy and roadmap on how we look at industry transformations which are happening especially manufacturing and then he obviously consults customers like you on what is the path forward and really answering you know some of these harder questions right so with that prashanth you know i'll hand it over to over to you we'll all sit back and listen in and we will come back you know with q a so over to you prashant thank you thank you very much for this beautiful introduction good morning good afternoon and good evening wherever you are in the world let's start today topic anomaly detection and diagnosis solution and first of all now before i go to the topic i would like all of you to understand why anomaly detection and diagnostic solution is critical for success of industry 4.0 or smart manufacturing and if you look at looking today world is facing three major problem and first is unplanned downtime and now when i say unplanned downtime is a which is the largest source of throughput losses in any industrial facility and when your equipments are not working reliably then definitely you are losing your efficiency also so and vice versa so if your equipment is not efficient then your downtime is more and all this leads to lower productivity so now no one likes surprises but it happens and in today's environment it happens more often so we will talk our in next few slides in next 20 minutes or so we'll talk what is our three-step approach to avoid surprises in a future and how you should plan a data driven operational excellence journey and we kept five minutes for question and answer so as i told you world is facing unplanned downtime and unplanned downtime is a reality unreliable operation reduce efficiency and it leads to lower productivity and this is the reason it is very critical to invest into the right technology which provides actionable insights and unfortunately traditional maintenance practices and approach are not good enough to provide you efficiency and to run your equipment at optimum up time and higher maintenance cost is a reality and 90 percent of time machine failures are independent of time so even though you have a very good time based maintenance of preventive maintenance strategy it doesn't help to avoid unplanned failures and to avoid those unplanned failures people are spending more on planned maintenance but sometimes access maintenance is waste of money and can increase chance of failure so now if you see the journey of maintenance in any industry now we have started with a reactive maintenance then we couple of years after we have invested in a preventive maintenance in fact so many people all of you might be also invested in the plc and kara system to monitor the equipment and processes but all these efforts are not good enough to avoid unplanned downtime or reducing cost of maintenance and one simple reason that alarms and alerts set in plc or scada are set for protection and it alerts when the already equipment is damaged so now next is predictive maintenance so if you see the journey in a maintenance reactive to preventive to predictive and providing predictive forecast and risk-based you really need technology to support you and provide you right kind of insight and there is a proof point in industry industry leaders who invested in digital technologies iot artificial intelligence much ahead of the time already started getting benefit they improved their availability almost by 8 to 10 percent they reduce reactive maintenance time by 10 to 40 percent and they improved labor productivity by 10 to 15 percent so now our iot by now real-time machine insight and foresight solution helps to reduce cost of maintenance increase uptime and labor productivity now our solution in nutshell will provide answer for three critical questions is my equipment healthy or abnormal what parameters contributing to abnormal behavior which part of the equipment fault is progressing so if you go into our solution we have a three-step approach for the anomaly detector and diagnostic journey so we start with collecting machine data and today fortunately sensors are become more affordable and you have more sensor than earlier but you are using sensor more for measurement and control but machine provide lot of information if you collect integrate and visualize and this is the most critical part of your anomaly detection and diagnostic journey second part is better detection so if you can detect early machine abnormally abnormality you can reduce your downtime you can reduce mean time to repair and because you have sufficient time to plan your maintenance activity and this also helps overall overall efficiency in the plant i would say fall diagnosis and prognostic is the goal in a journey better foresight and insight will provide a competitive age to the organization so first step in our solution is understanding as is analysis real-time monitoring of machine behavior in a various context of or in a context of various key performance indicator understand now what am i doing today versus yesterday what am i doing today versus last month whether my efficiency is improving or whether my efficiency is reducing today my all the machines are healthy my mtbf and mtta are kpi as per whole i o set or it is deviating so these are the most important contacts you would not would like to see your machine or assets and then anywhere you find there is a deviation you would like to analyze and identify root cause very quickly and you should able to see that which parameters is contributing to this particular fault in a machine so this you will get from our system very quickly and this help you to improve overall productivity and efficiency in operation and of course to avoid unplanned downtime forward using artificial intelligence and machine learning approach we provide alerts on abnormal machine state data science models monitor all the parameters 24 by 7 365 days it's similar to your best operator looking into all machine parameters every five identifying influencing parameters normal condition example in case when centrifugal pump in your water treatment facility detects abnormal vibration it also directs you to contribution parameters for particular fault it may be x axis rms value or bearing temperature high and this specific information helps you to prioritize and your work order next step in a journey is machine fault diagnostic using advanced patent recognition models in artificial intelligence and machine learning we can able to provide you specific fault mode detection so it may be abnormal vibration or you have a bearing wear or you might have a seal failure issue so this further provides you to prioritize your maintenance activity everyone in this world have having very scarce resources so your important is how easily you get accessible intelligence and then this helps you to prioritize your activity more efficiently so you can avoid avoid unplanned downtime productivity loss or production loss and how this solution work so we collect integrate sensor data from using our intelligent industrial iot age hardware so whether you have a vibration data whether your temperature energy pressure whether you have already sensor available or we can help you to new sensor connect to our intelligent industrial iot board either on a 485 or 232 or industry standard protocol like modbus modbus tcp or using the wi-fi or bluetooth sensors and maybe 4 to 20 milliampere 0 to 10 volt once data is available on this particular platform we we do the data quality checks using various algorithm then we run we do the transformation of those data and we generate alerts and we help you to see data in a context of subject matter expert on all various different contacts so maybe if you see this to screen you can see the machine status and drill down various actor to examine contributing parameters then you can see all your machine in various hierarchy drill down to the subsection either it's a mechanical or electrical and see various kpi and further drill down contributing parameter for that particular kpi so in last i want to tell you with my experience in this field see almost 12 years of field experience in industrial iot and overall 20 years i have seen this industrial automation journey and i would say industry 4.0 or smart manufacturing or another detection solution in the journey and first step is of course remote machine health monitoring collecting integrating transforming your machine data check data integrity and quality having a good descriptive analytics is a real real asset for your plan and then you can collaborate more real time with the various stakeholders and this next step is using how to use or how to mine those data so we have a artificial intelligence on machine learning based algorithm to help you detect your machine anomaly very early and then this give you a answer why is this happening and you can easily find it out what are the contribution contributing parameters and goal is basically machine fault board diagnostic so system should tell you that you have a bearing failure or your bearing failure is in progression and then you should start observing your equipment more closely and avoid those unplanned breakdown or you you can reduce maintenance cost or and you have enough time to plan your maintenance so uh this is what i would like to say for today's session and i'm open for uh q a that's that's great prashanth thanks thanks so much you know yeah forum you know absolutely fire away you know if you have any questions i would love to you know and to hear from you uh hi prashant nice session uh uh just wanted to see um you know like this animal detection like is it more like asset agnostic yeah of course so uh our solution is acid technology agnostic and we we first we try to understand now whatever sensors available on asset we will connect on our age platform and if we require we also can help you to add more sensor on that particular asset and help you to you know build the system for analytic detection and diagnostic whatever i said whatever make uh you have in your plan yeah so as a follow-up to that prashant actually i have a question so let's say if any of our customers they already have some kind of sensors deployed can can our scan your solution fit on on top of that seamlessly and i understand you know some stitching may be needed but you know yeah of course so the whole idea of our intelligent age platform is to connect uh existing sensors from machines yeah so it doesn't have to be a turnkey ground-up implementation right it can set yeah so most of yeah so reality in world is you know there is more brownfield opportunity than greenfield opportunity in the world and there is no there is more challenges in brownfield opportunity to connecting those sensor and that is where our experience of connecting asset and our we have designed our intelligent age platform to connect those available sensors uh or data equipment system onto our platform yeah thanks here any any other questions forum uh yeah prashanth so i would like to know can we detect these anomalies on real time on our machines yeah of course so the whole idea of our now if we can see the three-step approach in first step approach we we collect the sensors data in a real time okay so we we collect the sensor data if required some kind of vibration sensor or those even in a milli second five second or ten second interval and we provide you know basically solution to give you alerts real time it's similar like now one of your best operator looking into all your sensors every one minute or every five minute so that is what our solution uh really do for your plant oh yeah that's really awesome then we will have enough time to uh go with you know like uh with zero downtime uh unplanned downtime uh and i think uh this is this is uh what what the industry leaders are looking after okay so you will get enough time to understand what is happening with my asset and then you can prioritize your work you start observing more closely and you can find it out root cause and then whenever you get opportunity one thing one thing i am sure you agree that in manufacturing environment now opportunity to to get solve the problem is very less so it's very critical that in real time you understand what my asset is doing whether my upset is healthy or my asset already showing some kind of abnormality and my abnormality is progressing towards caution or warning level all right all right thanks thanks for awesome awesome awesome awesome guys so it's time to call it a wrap you know thanks again for for the time you know everybody who attended thanks prashant and uh thanks everyone yeah thank you thanks as a reminder guys this will be available on our website so you know by all means you know feel free to go look it up again share it with folks that that you need to and once again thank you thanks yeah that's fine thank you very much thanks thanks for your time today