 So, good evening all the newcomers please sit here in the front everybody is trying to grow back. So, as you your half the colleagues have already attended that so my task is to say the first sentence same and the last sentence same in between of course, I cannot promise that I will say the same thing. So, the whole idea I was trying to communicate with the your first batch is to what is the role of a effective English in defining the problem and therefore, I take the liberty of talking to C A students though it is not my field I am a electronic engineer and have done physics for most of my career experimental physics. So, I think you have to bear with some examples which are more common easy to understand for me and then you have to actually identify your problems in a similar way at a later stage because if I talk about programming and various other aspect of computer science and my knowledge is very very little. So, let us get on with a research problem we saw on Monday's lecture last that you know there is a five step process. So, define the problem then define the hypothesis plan the experiment make observations make analysis and see whether you achieved the target goal which is the hypothesis. So, hypothesis is like a PhD problem with your guide supposed to give, but invariably the guide gives you a topic and then he ask you to do what is called a literature survey and the literature survey is supposed to make a background ready background for you if you are a M Tech student it is for a M Tech one year project for a PhD student it is like a three to four year activity. So, let us say how effectively we can actually pose the problem and later fill in all the details which you get through this literature survey and various other coursework that you do. So, I will begin by two examples from what we know as the thinkers in the fields Stephen Covey you might have heard of his famous book of Seven Habits of Effective People and one of the important thing is begin with the end in mind. So, that means you have to begin your thesis with a wonderful PhD problem that is what it means you cannot begin at the beginning and as you grow you first start walking then you start slowly walking faster then you start running slowly and then you start running faster your target should be that you have to win the Olympic 100 meter goal right. So, that is that is called the begin with the end in mind. So, what is our end in mind my research problem should be a very good problem and if you should solve it there will be an achievement of some kind very similar thing was told by another group of authors Nadler and Hibino and their book is also breakthrough thinking and they say that whatever you are doing your solution should be having this property called uniqueness and the second one is called solve the next problem. So, let me at the outset say that in any PhD or any research there are only exactly two aspects first you have to define the problem at the end of your activity you get a solution or you get several solutions of which you claim one is a very good solution. So, you have a problem and a solution, but unfortunately for you you cannot do it on day 1 or beginning of the project because the human brain thinks in different ways. The left brain is full of ideas and the right brain is more of facts whether really you got a solution whether we can do this here whether we are enough resources here whether I have enough time here whether I can I will understand everything that is required. So, that actually limits you that is a fact finding side of the brain which actually looks at every solution whether you can do it or whether you will do it well and so on. So, I think at the time of defining the problem let us use only the left brain and let us divide this as when you use the left brain the assumption is that you cannot use the right brain and vice versa. So, that means for defining we should use only the left brain at that time we should not worry about the solution because I have already put a restriction axiomatic restriction that you can use only one side of the brain. So, this lecture is about using the left side of the brain we will not worry about a solution we will only talk about the problem how many types of problem can I have how will I actually very correctly write a problem in simple English because this course is a communication. So, I must be able to actually postulate a problem in very good English that you can understand I can understand and everybody will understand what you are trying to do. So, the PhD problem is like a research problem. So, pardon me if there are more than half the people m tech. So, you just assume that for the moment you are a PhD student does not matter. So, a research topic is good for both m tech and m tech. So, we will say research problem not a PhD problem. So, research problem so, where to create a mission statement or we what we call the goal. So, goal or objective is like I want to go to the moon and the scope of activities I have to you know make a launcher for taking me to that correct I have to make a rocket. So, that is the scope of activity the goal is that want to go to the moon. So, like that the mission statement is that there is a mission that I want to first determine the best research problem and once I get a fix on that then I will find several ways of finding solutions the good the bad the ugly and then we will find the right solution. But when we attempt to solve this we should have or even in the title of this PhD or research problem we should have this question what is novel in it what is the novel in the problem itself we are not yet found or you are not even thinking about the solution what is the novel or non obvious in the problem itself that means must be a striking problem. So, few of the examples which I will site later I will give you right now. So, I say that my PhD is on solar cooling. So, we know solar energy is heat and I am talking of cooling opposite or my problem is virtual reality. The reality is a reality and virtual is not a reality right, but I now combine this and I say my I am working on virtual reality or I am working on reconfigurable hardware hardware is supposed to be hardware right not supposed to change it. But now we say we want to change it configurable hardware and after several thesis is done on configurable hardware then of course where to invent another word called reconfigurable hardware. So, these are the kinds of titles that should impress us otherwise in my physics career I found most of the PhD problems were some aspects of ferroelectricity in this some aspects of something means a dudder you know already I reject it as anything novel because I do not even want to read it if you say some aspects of 3D printer, but 3D printer is good. Anybody heard of a 3D printer? Have you heard of 3D printer? I have decided not to ask questions of the first row people. I welcome them third row blue shirt have you heard of 3D printer? You, anybody has heard of 3D printer? Yeah please tell. So, I can give a drawing of a pen like this and ask the 3D printer to print little print a pen 3D. So, it is very difficult to imagine somebody who is familiar with a printer to imagine what is a 3D printer. So, that is why these titles are very very amazing titles and how it is a reality you can buy it for 599 dollars you will get the object of your desire printed. Secondly will there be that is another question a eureka moment I mean we do not want it to become a dull or dumb skip you know somewhere down the line got it. So, that means your problem is quite tough and having achieved a solution any solution is a eureka moment. If you give a very simple problem like some aspects of ferroelectricity if you get a result as good as anybody else you will be no thrill in that. So, I think your title is very very novel or as a mixture of this opposite sounding words then we are going to get a thrill out of it and then we will definitely expect a thrilling session with the research. So, begin with the end in mind you must always have that dream that dream is at the top of your you will win because I am telling you Olympic you may not be the 100 meter gold, but if you give a PhD title or research title you will solve it that is enough. See I have written every slide at the top of it it says begin with the end in mind that is it that the that the Dharma of research problem definition no where you forget the gold should be high reach even lower actually. So, let us try to make the best of our time available. So, let us say the end goal must be a top goal right. So, what are the goals like in a researcher's mind? I am going to write a list of goals my handwriting becomes a little smaller as I proceed that is why I have requested some people to come forward, but I am going to read it anyway. So, one to find a solution to a difficult or what we call or world's problem. So, your literature survey is not meant to find out what people have done and you also do similar thing. The literature survey suppose you are supposed to find what they have not done yet. So, to find out what they have not yet done which you should do to publish a paper in nature or science. What is so great about nature I mean it has got a very high index 20, 30 like that. So, what is the nature paper it is like this if something has been done in nature like every duck has got web fingers and it can swim in the pool in a freeway or every bird every sparrow has got only 10 grams of wings the feathers, but somehow the 10 gram support 150 gram of the weight of the sparrow it can fly. So, sometimes you wonder you know how you do that. So, that is the wonder. So, if you wondered about a tough problem and you find a solution this I have found a solution then you can go and publish in the nature because nature is all about solving some wonderful such phenomena. Then the next one is published many papers in top journals I would not ask you why you want to publish so many papers in top journal because the next line says. So, do a good thesis good enough thesis. So, that I will get a postdoc in MIT, Stanford, Berkeley, Cambridge etcetera right that is why we want to publish in top journals. If not this then of course finish in time because your scholarship drops to 0 or final get a job and of course it means bye bye to research. So, if you do not do anything about it is like a pendulum is going to fall down with gravity. This gravity is going to pull you down if you do not do anything about it I know people used to join TIFR with the hope of getting a Nobel Prize later they said at least one paper in nature some papers in IEEE or whatever is your association then a good thesis within quotes your guide says it is a good thesis or some nonsense like that. And then of course finish it in 6 years that was considered as good time because after that the scholarship is not there and finally get a good job. So, I used to tell myself as well as my colleagues younger colleagues later I do not know do something to stop this sliding down then I was so particular I kept on telling one fellow I thought he was bright. So, finally he reached Stanford that is good I mean somewhere it has to stop right otherwise it will slide down all the way. Then you wonder that why so many m tech students lose interest after December whether placement has been done and the job has nothing to do with their m tech thesis and they are lost. But I tell them I have been the job is already acquired, but why cannot you finish this project nicely the project is interesting might as well finish it is not it. Instead of lounging around in front of the TV you know there are mattresses there is that is fine, but it does not mean that in the last 4 months you lounge around there and do nothing. So, please do something about it so that there is no downward trend. So, end goal is a top. So, what do you do in let us say the initial period what we call the litigious survey. So, what you do is you read everything sorry for drawing this brick wall opposite to the way it is constructed during construction it goes from bottom to top, but this is the way we do litigious survey we read papers in 2012 first and then in 2009 then in 2001 and maybe 1995 we do the litigious survey in the reverse which is ok. So, what we do there is you try to find a gap they say ok my object is to publish 2 papers here. So, people have left this gaps and fill up the gaps, but that is not going to agree with my original definition of defining your research problem as a top top goal or a mission. So, we say we will do it we will find some excuse for doing this also because that is interesting part is that human mind is a wonderful phenomena by the God that he find justification or gratification in doing anything and everything. So, I keep on telling you must try to work for a noble prize or a fields medal or whatever it is, but I will also like you to publish some full 2 papers also you know I am done. So, I will tell you how to fill up this gaps and publish paper, but that is not the goal your job is to make a new structure in which some other stupid people will fill up the gaps you know. So, you have to now make a structure which is new and that is that is the whole business about the top goal. The research goal is make a new structure. So, I am told for example in computer science the only great thing that has been done in last 20 years that is what I am told is what is called the design patterns. I am going to in another 10 minutes I am going to give you this pairs of words which are very interesting like 3D printer and virtual reality and solar cooling you know exactly opposite meaning. So, you see any opposite meaning here the design means you have got the freedom to do it whichever way you want, but the patterns means there are fixed number of patterns. So, this gang of 4 found some patterns and they have made a list of patterns and they now say there is nothing more to think about these 17 patterns or whatever that number is that is all is required. So, that is called top research because now you are defining taken away the thrill of doing new designs because you must now use their 17 or whatever that number of patterns is and then all programs will be written perfectly. So, I think there is something that everybody does you have to move exactly in opposite direction and kill that the routine activity what I call. So, what we call this called a routine is the routine research and we solve this contradictory things are going in opposite action and you find that conflict and solve that conflict you are away from the routine. I am very happy to hear that tune well before the birth of your father may be such a old song from Nayador. So, now we are back there we always have this top that begin with the end in mind. So, that there is our title for the day begin with the end in mind. So, question one when we are at it defining our problem question one are you solving the wrong problem. So, that is the biggest worry you may do all of the work and then in the end somebody will say but that was not the right problem to solve because there are things called right problem and wrong problem and therefore, I am going to define what is the wrong problem and therefore, opposite to that is the right problem. So, take an example. So, I do consultancy with Mahendra Mahendra tractors. So, there is a problem called tractor bonnet which is painted nice and red you know you have seen those tractors as corrosion over time. So, you get corroded and you suddenly start seeing some rusty patches on that and then it spoils the whole show. So, they have taken this task to find out what is the cause of rate and then solve it. So, the sequence of events is like this one they find that the during electrochemical deposition because the electrochemical deposition is the best way you can do it. So, you have this primer on the steel part you put a primer on which you are going to put the pen the primer is done by electrochemical deposition. So, as soon as the primer as I say coat of 1 micron or 2 microns it becomes insulating and wherever there is a pinhole that means the bare metal is visible to the electrolyte the electrochemical deposition occurs there. So, now you see this wonderful thing that it itself finds where there is a bare metal and fills it up plugs it up. So, that means at the end of this ECD deposition there will be no bare metal by definition right. The current will automatically die down to a particular low value and that shows that now we have got a perfect primer deposition that is very good, but there is a bigger problem because of the shape of the bonnet like that and then there is a tackle with which you hold it and this is the electrolyte and the electrolyte is slowly filling up there is air bubble trapped air. Now, you have got a trapped air and of course the electro deposition will not occur there and then that is the point where it will start the problem. The trapped air could be anything which has got a small suppose you have got a small welding or something like this then there will be a trapped air here also trapped air. So, anything you give for holding that bonnet down finally with I know that rubber clamp or whatever it is there is a possibility of air being trapped. So, what they do is for this trapped hole trapped air and this because of the shape of the bonnet you can do nothing about it. So, what you do is you insert a tube and this is the productions of the tube is also of metal steel insert a tube and remove this air. So, the problem has been fixed, but then the tube gets plated with the primer and the tube being tube which is very small slowly clogs and then you have to clean it. So, third one is the tube gets coated get coats and therefore, clogs. So, now we have to have a brush at the end of every shift you have to have brush it and clean it and the brush has to be strong because it has to remove the primer and then again do that plating again. So, there was no such brush because this tube had a this tube was coming and had a bend like that a nice natural bend. So, you have to have a very flexible, but steel like springy wire which can bend like this and has have bristles on it. So, the solution to that problem was design a new brush with spring steel wires and the very hard material bristles and then they solve or they are solving that problem and then when I went there I said oh it is not the right problem the wrong problem I said what is wrong with that I say it is doing nothing with the original problem. So, that means the guideline therefore, if you agree that it is the wrong problem to solve the original problem is miles away. If somebody is designing a brush you will not even get an idea why what was the original problem forgotten and therefore, we always say that you solve you must solve the original problem. So, our whole stake is on this word original that means right in the beginning first 6 months you must postulate your problem and must stick to it on the way to the solution you may find some secondary problems or solutions having drawbacks etcetera we cannot do anything. So, this tube codes and clock this is called a drawback of the solution that particular solution by which the air was being taken. So, my suggestion to them is make a small hole 1 millimeter hole or several 1 millimeter hole and finally, when you paint it you can close that hole it is too small to be seen and so on yes and no design change I said that putting tube also is a design change. So, they hide behind this extraneous criteria that it is a design change and so on. So, you might also do the same instead of solving the right problem you will solve the wrong problem in a detail and somewhere in the third year one of the internal exam you will say do not you think professor that we have to change the title the title of the this is naturally is not it we have to change the title of this is. So, you are very meticulously trying to solve that problem you forgot on the original problem. So, let us not do that. So, let us define the tough problem and stick with it or even if you solve intermediate difficulty you must keep going back keep going back. So, my advice is always solve the original problem and not the drawbacks of a particular solution which you will also get involved. So, original problem was how to eliminate the trap air and the their ways of eliminating trap air my better suggestion I think was make a few holes at the critical points and later on close those holes with the paint itself. So, now we want to use good English to define a tough problem. So, what is the meaning of good English the good English in is only noun verb and noun all the adjectives and proverbs and all that and any clauses we shun because they actually dilute or divert the problem. So, we are going to define a problem in that simplicity of a noun verb and a noun. So, there are three such jobs in simple English very very simply I have to say the air is causing corrosion air causes ultimately corrosion the road erodes the tire. So, as simply as that problem should be mention in that way. So, one is mention your problem there are three ways problem in as a conflict as a conflict in two desirable parameters called a and b. So, I have a problem that I have a huge program with hundreds of modules and I have to integrate with so many other modules at the user end and so many other modules at the final customer end and then I am in the middle of it. So, first is that I have to be sure that all my modules talk to each other properly then the front module which interact with the customer entry point also talks nicely and finally I give the end result the way the user wants finally. So, one is my own modules and the other one is interface with everybody else let us say both are desirable, but what happens is the more I do on this this becomes poor the more modules and more things I do more likely than not that it becomes difficult to integrate with anybody else then I say ok let us keep the integrability as a standard objective and then I get a limitation on my my own module because my module must exactly have that format etcetera. So, I think we have to solve this problem perennial problem forever. So, you have a huge big program like SAP and then it must be Niap is a perfect right and in billions of dollars or trillions of dollars of things are going through that program and still must not give an error of even one cent. So, that is the difficulty and not a single customer should be unhappy. So, that become the a big challenge and therefore, in my problem you define what are the two conflicting requirements if I do something what I do improve B then A has a problem if I improve A then B has a problem. So, one of the classic problems in microprocessor is if you increase the speed right gigahertz and you increase the number of cores in a in a package 4 core now we get a 4 core microprocessor. So, that is the good we say A is good. So, increasing the speed A is good, but then the power dissipation is a problem I do not have a red pen you have to bear with it. So, what is the VLSI designer doing who is designing the chip is just putting more and more and more components of the chip and the speed is increasing and half c v square f. So, f is the frequency and c is bigger if chip is bigger half c v square f. So, now, we got to reduce the voltage, but you cannot do it indefinitely. So, you got this power dissipation even 10 years ago that Pentium 100 megahertz used to have a fan sitting on that you remember inside that main processor chip had a separate fan inside there and now of course, it was only 10 or 20 watt now it is more than 200 watts and very soon it will become 1 kilowatt. So, now you have got a bomb inside your microprocessor the bomb is inside your laptop. So, now you have got a cooling jacket for laptop. So, how are you going to solve this problem it is not a electrical engineering problem it is a heat transfer problem. So, one of my favorite student from mechanical engineering joined Stanford University Integrated Circuit Lab and he was working on the heat removal from this integrated circuit chips because it is a tough problem. So, you do not want to change your design activity you cannot do that easily. So, you hire somebody who can solve the heat dissipation problem he is the champion. So, invariably in any technology the focus of the current research actually moves from one field to another field to another field and you cannot keep on saying that the computer science is one such thing with a fixed idea. The focus of that activity can go to communication between computer or getting utilization of this computing power in the world through cloud computing or like. So, this focus keeps on changing and therefore, you have to know where the focus is where the crunch is. So, you have to say this is important I have got to do something with that. Second one is that software must be easy to use this is a very favorite example a 4 year old can work on it and it must have complex features. So, that is windows for you a 4 year old can use it just like 60 year old like me also can use it I spend 20 hours I become the computer why I do not understand what it is right. You may drive a car for all your life and may not know how the engine works. So, it is like that we are using it only as a end product we have got pen drive we do not have to know what is inside there. So, we keep on using it, but then it is affected by virus. So, we have got line as we say is not affected by virus, but then nobody can use it neither 4 year old nor 60 year old. I do not even know the first sentence a first instruction I have to type to get on the line x. So, now ultimately what will happen the line x and windows will all appear like a same same guard right the wonderful software which anybody can use will have features of both combined. Finally, bottom line therefore is my solution must have a good and b good it is not or it is original problem appeared as if a or b you can have high speed or less power, but it is always a and b I want high speed as well as a high power which I know how to remove take example of super computer. So, I was with Tata who started a super computer in 2006 7 and there was the fourth fastest at that time 120 Terahertz. What is in a super computer I do not know, but I knew there are 1500 processors sitting there and consuming power like hell and if you have got a bigger super computer now petaflop the flop flop is not a English word called flop the petaflops are flip flop like how many times it flips and flop 10 to 15 times per second it consumes 12 megawatt. So, let me be require a power generating station right next to it is a hall like this size 12 megawatt energy is being consumed in this hall. So, first is how to distribute energy 12 megawatt in such a small area and even bigger challenge is how to remove it. So, give a contact to the Tata company another Tata company for cooling it it was a horrendous job first to dissipate the power cables have to give and give that power and immediately remove it. So, double job over here for dissipated and then remove it remove the heat. So, now we focus on removing the heat. So, they have what a this is the computer room triple this size another room was required for the heat exchangers and all that cooling apparatus like the full basement in our lecture theater you can go down you know the big cooling system there you know making awful noise there. So, three times that size is the cooling apparatus required. So, if we take a a view that removing the heat, but not having such conventional cooling then I will say cool it by liquid nitrogen. So, why should I have a liquid nitrogen plant I have a liquid nitrogen which is excellent cooler. So, I just get a tank every day and then that one tank is enough for cooling it several hours or whatever why put a 12 watt generating station we take it from a separate feeder from Tata's or reliance. So, like that you must avoid the unnecessary burden of solving the wrong problem you know installing this power station is a wrong problem and removing the heat by using conventional cooling voltage like is also a wrong problem. So, you must have the right solution get a feeder from the suppliers and use liquid nitrogen and liquid nitrogen is not expensive because we unnecessarily think. So, it is cheaper than the milk per liter. So, milk cost is some sixteen rupee per half liter or thirty rupees a liter it will be costing only thirty rupees a liter, but it will do a wonderful job in removing the and then your super computer will work more efficiently because you are not unnecessarily burden yourself with the wrong problem otherwise you will be solving the wrong problem forever. So, we want a and b and not the burden of any one of them a and b always. So, the conflict is either a or b, but we want a solution which is a and b, but today I am not trying to find the solution I am only postulating the problem. So, second problem type if I have a adjustable parameter let us say c then it happens like this suppose c is a parameter at some high value it gives me some good function a good. For example, if I am cleaning my job to be painted which is the bonnet of my tractor then cleaning is done by spraying some acid five percent acid. So, you keep the concentration high it cleans well. So, concentration of the acid it cleans well, but then there is a problem pollution because once you have this spray and it bounces of this bonnet and then creates a whole lot of mist and then you have to have a special action for sucking it out then you have to have a scrubber then you have to have a actual NOH mix it with NOH to neutralize this acid and all these requirements are coming why are they coming why are they not there till yesterday because now Mahindra and Mahindra want to be a world leader if you want to be a world leader you have to do all this if you are sitting in China you do not want to be a world leader then you can do anything in there you can dump this acid in the drainage and it will take care of itself that is what you think, but when you want to really solve the right problem then pollution B is bad. So, if we keep the concentration low for some reason I can even use 0.1 percent acid and get it properly cleaned then the pollution is ok pollution problem that is no pollution good, but then cleaning itself is not good cleaning poor. So, A bad we have got A good and then we have got B bad and then we have got B good and then A bad, but what we want is both A and B good. So, you know what is the solution in that spray gun you put electrostatic discharge and actuate this with a dry nitrogen electrostatic discharge and this electrostatic discharge goes with the acid and create those independent ions or what you call the charges of this and then they do much better cleaning. So, you can get it done with 0.1 percent HCl. So, only the gun has to be modified. So, that is the right solution what is the right solution I must do it with this. So, I do not have to do this scrubber and everything this is fine. So, that is the right problem otherwise you could have taken this and that would have been the wrong problem because then you do this you have to remove it by exhaust then you have to do scrubber action dissolve all that in treat with NOH till it is neutralized and then send that outside. So, you should not unnecessarily do the long drawn wrong problem correct solution business. So, among these we have got two choices one direction is the right problem. So, therefore, in the right problem you have identified and if I now remove everything just keep this. So, I remove the jargon. So, no jargon meaning no detail just the English right this is just the English there is a parameter at one extreme a is good but b is bad other extreme b is good but a is bad and now you have to find out which one is the right problem because we ultimately want both the end be good. So, this one is called mother of conflict. So, this was two sorry. So, why is it mother of conflict? So, c is the mother and a and b were the two children which was the step one right conflict between a and b. If I want a very high speed then I have to put 12 megawatt power in and b was that I have to again cool it. So, a and b were the speed of the super computer and the b was a cooling problem. So, there I did not say what was the root cause of it. So, now I say that c c is the mother has got two children and now you have to do best by both and therefore, this is called the mother of conflict. It gives a wonderful direction to your thesis not only you know which problem you have to solve a good b good you also know where you begin you begin at c because c is the origin from where you can begin your solution. But I am not thinking about solution I am just saying that it is a very good problem definition because c also is a direction towards solution. So, next lecture you wake up next day you will think about solution today we will only think about problem. Yeah. Yeah, yeah exactly. No that is ok, but as soon as I see such a diagram one of the two must be a right problem. So, let me repeat it for the sake of recording. As we correctly say that because for that specific example it was clear that this was a good solution. So, now in the hand side we can say this is the mother of conflict, but what I am saying is there is a good diagram which has all the necessary detail requirement of a good problem definition. I am not saying whether low is good or high is good one of them will be ok. One of there is a right direction one of them is a wrong direction that is all I am saying. See I have to give some example otherwise you will not even understand what all this happening. So, the goodness of this is that if there is a big conflict if I also have a inkling as to what is that the base of it then I have a possible direction to begin my activity. The third one is even more important this called elimination eliminate the harm. So, we see heat generated in that super computer as a harm right or the pollution generated in that cleaning as a harm. The fly which has fallen in the tea cup is a harm like that there is something good very clearly you know what is a harm. And therefore, it is very easy for us to simply make a statement that this is the harm and I must eliminate it. So, the heating whatever is happening I must find the quickest way cheapest way to eliminate that heat from the system because the heat I require to get the speed I need to spend 12 megawatt, but heat is the harmful temperature rises that there is a problem like that. So, same problem I can look at it from all these three ways that is what I am saying the same problem. So, when we say eliminate the harm it is like this the diagram is graphical. So, I give some energy to a thing called tool and this is my object. So, it acts on object or hammer breaks rock the super computer individual processors consume energy and therefore, create heat. So, electrical energy given to the super computer that unit core or whatever you call it creates heat when it goes to the final destination, but let us not dwell on that the problem is along with this good activity there is a harm. So, now, there are exactly three ways of eliminating this harm mentally. So, first is find the root cause and eliminate it and eliminate it. So, many times the root cause is very very difficult to find for example, when you break a rock with a hammer even how the rock get broken by hammer hits then analysis of that will depend on the place from where the rock has come and the structure of the rock etcetera etcetera and it is a very complex problem, but if you can find it you can do that. The second is find the action itself which causes the harm, but invariable the action actually is required to create the useful activity. For example, I have to supply that much energy to super computer to get my speed and therefore, even if I know this is the action I cannot eliminate it I need that. The third one very interestingly is a eliminate only the effect of the harm. The heat is generated 12 megawatt is converted to heat we keep on saying I do not want the heat I do not want the heat that not enough. So, we must ask the next question what is the heat going to do it is going to increase the temperature. If the temperature goes beyond 100 or whatever is the specification the computer will shut itself off rather it will get damaged. So, usually it used to happen in our initial you know early teaching problems were these that before we know 1 or 2 or this you know individual course will stop working where they getting heated then we have to worry about why the cooling is not efficient there. So, the temperature rise due to the heat is the problem right. So, the effect of the harm in that case is the rise in temperature and therefore, we must focus on the heat removal. Therefore, the diagram now I draw another diagram actually it can be drawn just below it. So, the rest of the thing remains same because I want the useful action which is useful action is my tool is my object. Now, to show that I have not eliminated the harm I still draw this wiggle half way. So, this is a very good graphical representation of the problem and then I have something some other substance or some other activity which will now on every activity like this sorry for the space crunch I have to draw it again here. The tool object is a useful action the harm is still there, but now I have got this some other material and the verbs have to be here on this. So, what does it do? Absorb or isolate or divert the harm. So, that is what I am just going to take away the heat out and therefore, this becomes my liquid nitrogen in this case because I want a ready meticulous I do not want the whole air conditioning system which occupies 3 times more space than this and takes a lot of footprint and it is a headache to maintain it. So, I keep the headache of giving liquid nitrogen to someone else, same way I keep the headache of providing 12 mega watts to somebody else and I do not want to solve that problem. See, if I am having a problem with some metal pipe joint is leaking or has got porous I do not go and start melting that and making pipes and all that right I have to find the right vendor for that who gives me pipe which is you know free of holes or voids so like that. So, nothing to be done on power side I do not want to solve that wrong problem I do not want to solve this wrong problem I want to solve it by using ready made coolant I am just saying as an example and therefore, this diagram the good action has got a verb on it. The absorb isolate divert this action has also got a verb on it and therefore, liquid nitrogen diverts the heat right this is the heat that is it liquid nitrogen diverts heat unfortunately I have gone into the solution mode. So, let us remove this for the next time we are stuck with this problem here. So, when you define the problem you should simply draw this diagram and be happy they have defined the correct problem I must get rid of the harm, but the good English is I must get rid of the effect of the harm not get rid of the harm only because harm will imagine causes me to change my technology what we call you have to re synthesize the whole thing. Somebody someday IBM I do not know somebody someday will make this super computer by some other technology which uses some actor joule for every flip of 1 and 0 right now it is giving me so much thing because every flipping is consuming some energy. So, somebody goes through orders of order down for every flip that is. So, they say optical computing may actually do that reduce that energy for every 1 flip 3, 2, 6 orders of magnitude lower and then I will not generate this heat. So, that means you have to re-synthesize the whole thing to get the harm eliminated are you willing to do that re-synthesize the whole thing make the super computer in a totally different way on new technology that is too much in the future, but if you have to solve this problem you have to only eliminate the effect of the. So, that is the good English eliminate the effect of the harm. So, if I get a headache do I go to a sorry have you got any other lecture I will just summarize because I do not want to unnecessarily over shoot I will summarize now I have got all this thing I will just make a soft copy and upload with the help of prosa area and yeah. So, looking ahead there are 5 levels of solution. So, I am going to discuss in my next lecture. So, first level level 1 is called apparent solution and it forms 30 percent of the pattern database. So, 30 percent of the patterns based on a single person knowledge once he specifies the correct problem he gets a pattern the right problem. The level 2 is minor modification 55 percent. So, putting instead of one core putting 2 cores and 4 cores as far as this method is concerned is called a minor modification. For Intel it may be a major modification, but for a user it is just a minor thing instead of 1 core now you have got a 2 core 4 core. Because you have got a already a existing good process you just put 2 of them because you have the my nanometer sizes are reducing. So, 65 nanometer node has now gone to 32 the 32 will go down to for the smaller so you can put 4 of them. So, that is called minor modification level 3 is called major modification that is only 11 to 12 percent. The fourth is very important is called paradigm shift what is a paradigm shift for supercomputer are we going to multiply 2 numbers of 60 h each or things like that forever. No we can do quantum computing totally different thing you can do optical computing of course doing the multiplication exactly the conventional way, but use the optical computer for doing that, but quantum computing is totally different. So, that is called paradigm shift. The level 5 is new discovery. So, you may yet win a Nobel prize for this. So, the fellow invented of made the integrate circuit first in 1968 what a Nobel prize in 2000. So, in TIFR they used to say not only you have to do great work you have to have long life you know you have to be alive to get it because they may give to you after 40 50 years. So, what is the significance of such solutions? This solution goes to conference papers, this goes to top journals, this may go into may be this or this will go into nature or science or whatever you think is wonderful this one nobody knows we cannot prepare for that. So, I think summarizing you have to use good English to define the right problem particularly in the case of harm elimination the good English is eliminate the effect of the harm not the harm itself. And several of many solutions and I think in your entire research career for PhDs all these will be valid as you progress for m tech student I think you must aim at one each a six months work must get you this two publications you deserve it the guide wants it, but you deserve it all that you have to do is define the right problem right at the beginning. And when should you finish this work before November before the placement begins after the placement nobody does anything. So, before that itself in the summer itself you should finish or some of you in the second semester of m tech right. So, define please define the problem before the semester ends and solve it in the summer vacation that is it. If you define the right problem solve it in the summer vacation everything is achieved you deserve this two papers and the guide wants it. So, he also thank you very much one more session on how to having defined a tough problem how to solve it.