 I propose to take an example of a particular information system and to describe some aspects of how information is handled. The example is that of an industrial organization. Typically, an industrial organization manufactures and sells goods and it has generally the following broad functions that have been listed here. Which are these, manufacturing, which happens in plants or works, marketing, selling, then accounting and finance when you collect money. Material management because when you are manufacturing something, you are purchasing raw material from some other people and you have to manage the inventory of raw material and finished goods. Then administration, human resource management, you employ people, you have to re-skill them, you have to train them, orient them, you have to handle their leave, salary, whatever. And finally, the corporate planning. Whether you should increase your manufacturing, reduce it based on demands, etc., etc. This is the strategy. Of all these, we will look at the inventory management problem. Most of you are familiar with the word inventory. Anybody here who does not know exactly what inventory is. Because industrial engineering and war people or such people will know that. Inventory is defined not, hypothetically is often defined as a cost, as a value, rupees. And it comes out of, let's take raw material and sub-assemblies which are needed by some plants for production. So, a material manager proactively procures material and sucks it. So, let's say you are manufacturing fans. And you have, let's say some coils or some rotor or blades which have to be ordered from sub-assembly providers. You have to keep them. Ideally, if you are going to produce 200 fans today, in the morning somebody should supply all the parts of those 200 fans and you should assemble and sell them. And actually sell them and dispatch them. But can this happen in reality? No, that fellow may not supply. So, you have people who are employed to produce fans but they have no raw material. Or you produce 200 fans today, 200 tomorrow, 200 day after tomorrow but nobody is picking them up. So, you have to store them. Every time you store either the finished goods or raw material it represents some cost. Because you have paid money but you are not able to exploit them. That is why a stores clerk who issues material against request and receives material from suppliers has to manage this inventory. And the cost of such stored material till it is used is actually the inventory. So, let's say I manufacture 200 pieces of equipment for which I buy say 500 different items from different resources and store them for different times. Then let's say at any point in time the average cost of all the items which are pending or stored is said 2 crore rupees. My inventory is said to be 2 crores. Why would I like to reduce the inventory? Obvious reasons. My 2 crore rupees are sunk there because I have already paid. And unless I recover money by selling the finished goods that money is blocked. The interest on 2 crore rupees plus the cost of the remises where I store this is all costing. I would like to minimize this inventory. Ideal inventory should be zero. It's called need to require basis. Just in time inventory as it says. As I require for manufacturing I should get it. This does not happen normally. Please note here the important point from an information system perspective is the stores clerk. Stores clerk is expected to get request for material. Get me 30 rotas today because I need them. Get me 2 coys for some equipment. Get me 20 fan belts. This request will come from production plant. The source clerk will issue those and record. Occasionally he will receive things which the supplier has supplied. So you will have to make those entries. That is the focus of the information system that we are talking about. But we are talking about this in the context of inventory control. Here is an example of a simple inventory control. The cost of inventory as I said comprises of price paid for the goods, storage costs, interest burden and cost of the procurement process. There is no free lunch. If I have to purchase let's say 500 fan belts then I have to ask 3 or 4 vendors. I have to write letters to them. My salary is being consumed. My time is being consumed. All that adds to procurement cost. Lead time for procurement. What is the average time by which I will get that thing? If the lead time is 10 days then I would order roughly when about 11 days of stock is there so that I don't waste my money in storing unnecessary things. But if I goof up my production will halt. So it's like a tightrope walk actually for the inventory control mechanism. Here is a simple example of an inventory control. This is my home. The morning tea is made by using sugar from this small dabbha. Many of you would be familiar with such stuff. Typically every household in India has this. This is a small dabbha. The sugar will be consumed here in 3 days, 4 days, 5 days, 10 days depending upon the usage. So what happens when that event takes place? Typically my wife will go to a bigger dabbha and take out sugar from there in this smaller dabbha. I agree that the process generally you are all familiar with this. While doing so, every time she will notice how much sugar is there in this bigger dabbha. When that bigger dabbha also gets near exhaustion she will say, now the new sugar should come. Notice what she is doing in the process. She is deciding when to buy. She does not buy only when this sugar and this sugar both is exhausted because she knows about the lead time of the sugar purchase. What is the lead time? The lead time depends upon how lazy her husband or son is, how long he or she will take to go to the market and buy it. What is the availability of the sugar? What is the availability of cheap sugar in which shop? And she will know that she will have to spend 2-3 days. And she will ensure that last time she fills up this small dabbha she will say, that is her inventory count. This simple system by the way is called a two bin system of inventory control. Those of you who have handled inventory will know that this small bin is for normal daily use, larger bin is for regular stock. Last time the small bin is loaded more sugar is ordered to refill the larger bin. This point is called a reorder point. That means that is the point when you reorder a stock of inventory and you don't reorder small dabbha worth. You reorder 2 kgs, 3 kgs, whatever will go into that big dabbha. But at what point you reorder is important from an inventory perspective. You call it reorder point. At that point you reorder. How much you reorder? You don't reorder the full large dabbha because some sugar will be there always. So you reorder a certain quantity which is called reorder quantity. The reorder quantity would be ordered on a certain date and the beauty of the inventory system is by the time the supplier supplies the reordered quantity the small quantity will be sufficient for your normal production. That's the fundamental of this simple scheme. In any industrial organization there will be variations of this and several other models because it's a very complex problem. Imagine you have 10,000 items in the inventory. Each item has a different set of delays associated with supply dates. Different pricing is associated, different costs are associated. So it's a very complex problem. I'm just simplified. In this simplified mechanism what are the material related activities that happen? Please note we are trying to answer questions of what information is needed, why it is needed, when it is needed, by whom it is needed and we are making the store's clerk to be our focal point. The store's clerk is sitting in a store. He has multiple bins in which items are kept. The plants will keep sending requests saying give me 10 of these, 20 of these, 15 of these and they will receive internal material which is issued by the store's clerk. It is obvious that any time a reorder point is reached the material clause must send a shortage report to the material management system and the material management must order the reorder quantity for that item with the supplier and eventually the supplier will supply that material which will have to be accumulated in the store. So you will see the information that is required to flow, request loads have to flow here. The actual material will be issued here but the recording must be made here. The obligation must be such that shortage, whenever it happens a shortage report must go there, a purchase order must go there. Supply of new material which comes in must be accompanied by vouchers or purchased order date or whatever, whatever price, whatever. So this is the kind of information that needs to flow. How does store's clerk maintain information of all these things? This is what we are trying to see here. So the store's clerk must first verify the authenticity of the request, must locate and issue material and record data for each transaction, must check if balance is less than the order quantity, if so inform the material management and if there is any receipt of material because the previous order's clerk verify and record supply particulars and send information to material management and to account saying I have received this. Typically he would use a laser in a paper-based system, one page per item in the laser. The top part of the page will have overall information about the item. Let's say part code so and so, it's a small fan belt, its storage location has been number so and so in thousand bits that I have. The opening balance at the beginning of the year was 250 fan belts and the reorder point is 50. At any time I have 50 or less, I have to ask for reorder. Such a thing will be handled in addition, remaining page he will write the transaction history. So there is one page in the register. In that page top portion describes the material characteristics and the last portion describes individual terms. So let's say 14, 12, 0, 6, plant A asked for 25 items. At that point the balance became 130. 250 might have been the opening balance at the beginning of the year. I am just showing sample transactions here. Another date 45 items were asked by plant B. Another date 38 items were asked by plant A. Notice that this 47 represents reduction of the level below 50. 50 was the reorder point. This is the point at which that means 29, 12, the clerk must inform the material management of. Notice that there is another transaction request on 2nd January 2007, but that is okay because I still have material. Typically then the page in the stock register will look like this. So top page, I have hypothetically called it Power Electricals. The top page describes part code, part name, storage, location, etc. and the last part describes all the transactions. Please note that these transaction entries have occurred here based on the request notes that the plants have sent or alternately the invoices which are supplied by the supplier. Note that this entry is completely different. Some say CXVP said 31, 12. This may mean a purchase order given on 31, 12, because 29, 12 is when I reported that there was a shortage and this says 350, but here instead of quantity reducing it has increased from 17. Very clearly this is a supply of material that has happened. This is a simple and elegant scheme and many people follow this scheme. Now imagine you want to permit the stores clerk to use a computerised system to handle this. The stores clerk will have actual physical material around him but he will have a computer. So he will get request notes. Let's say he gets request notes still on paper, but you would like the records to be maintained, managed through a computer programme. If you have to write a programme, how would you go about doing it? It's not a very complex programme. All of you have done programming of this kind earlier or the other but can you think of designing a programme logically from the perspective of identifying not just the information but the way the information will be represented, the way computations should be made, et cetera, et cetera. That's a question. I mentioned the previous page if you look at, you notice there is a page number, 138. If somebody has 1000 items there will be 1000 pages. Suppose a request comes for fan belt small, how do you go to page number 138? Would you expect the stores clerk to remember? No. So there will be an index. Just like the key issue register with the watchman here has an index page at the beginning. So you will have an index page which will say part code this, page number this, page number this. Sometimes that page gets over and you are provided only one page. There are more transactions. Then the fellow will introduce a new page, copy the top information, start writing transactions there but he will write a new page here. Please note that without the use of this index it will be impossible for the stores clerk to handle the requisition quickly. Essentially then any backend storage that you plan must permit storage of the master information about part, must permit storage for individual transactions and must permit building of an index and its use in quickly locating the record in the file for that particular transaction. You can either do batch processing of all the transaction requests. What does it mean? For each requisition slip that the stores clerk receives issue the material and put the slip in a transaction file. Just keep storing it in a file. At the end of the day, update the appropriate item records in the ledger. This will cause delay in updating of information but handling of the transaction will be very quick. You are not wasting that fellow's time waiting for the material. On the other hand, you can do online processing. As each item is issued, search for the ledger page, update the quantity immediately. There are pros and cons of both approaches. Batch processing avoids delays in main activity that is issuing the material but delays information up. Consider submission of registration forms in the old style type. So all of you submit the registration forms. They are all accumulated by the academic office and over one week they enter the data but the delay will be in my getting the course list. On the other hand, an online processing where you submit your registration immediately and automatically the course list is being formed as you enter the registration. At any point I have the updated course list available to me. So obviously the OLTP is better. The operational alternatives for this activity is either a card X which is called a card index system. You may still find in small and medium sector units. They don't maintain computerized information system so they will have a card X, one card per part. And you will have cards which are stored on part code so that you can quickly access the card. Or an automated system which is programmed on a computer. Or you might have a very comprehensive information system implementing all aspects of accounting, payroll, inventory control, sales, marketing which is called an enterprise resource planning package, ERP package of which the inventory management will be just one part which will also be handled when you implement an ERP system. The desired OLTP or online transaction processing functionality for the store's clerk can be documented as follows. A, maintain a master file for part information. Now when you say master file you have to define a master record in that file. That record will contain so many fields part code, name, stock, etc. You collect data for a proposed transaction through a screen so you will collect a transaction record. There will be master record for all parts. There will be a transaction record which is a part code type of transaction. What do you mean by type of transaction? Issue or receipt. The quantity involved. There may be other information. This is just the bare minimum information that you require. And for each transaction the activity that you will do is read the master record for that part. Depending on whether it is issue or supply you will add this quantity or subtract this quantity from stock and rewrite the master record. Simple activity for a programming is not very difficult. But as I said, professionals don't start writing programs. They design. So far what we saw was the analysis part of the information. Now we design the program. So designing means you define very clearly what is the input to the program. So part code transaction type quantity of transaction. What is the master record? Part code, part name, reorder point, reorder quantity. The transaction log information will be designed. For every transaction you will create a log record in some other file saying this is what happened. This is important for non-repudiation later or this is important for finding out all the activities that have happened. And then exception report line. Stock falling below the reorder point at any time itself. You will have part code, name, quantity, whatever. You will then design sample record structure. So you will say a file on the computer will have a name inventory master file or enumast file. You will define a record name for that file which is a variable, which is a string variable let's say. You may call inventory master record. You will define the fields if that record. So first field may be part code which will be six characters long. Next field may be part name which will be 20 characters long. Next field may be reorder point which may be four digits long. Next may be quantity which is numeric five. Num four, n five, et cetera are different ways of stating the same thing. Incidentally if you ever have a brush with cobalt programming you will find that this is how you will design the record layout, the various fields in the record, describing whatever context you have to decide. And the sample records in master file may look like this. Notice the way the record looks on a computerized filing system. FB 1376 SMALR. There is no distinction. Why? You already defined first six characters are part code. Next 20 characters are part name. Next four digits are reorder point 0050. Although this number appears to be a single number it is not like that. These are different fields. There is no field separator required in computerized systems where the metadata is very clearly defined as how many digits mean what, how many digits means what, et cetera. Similarly a sample record in transaction file may look like this. So FB 1375. This is what you call part code. This may be transaction type R. What could be the transaction type R received? 0350 quantity. So such could be the design that you will do. And then you will write a program. I am not going through a program although I will try and put on the web model for this course a cobalt program for you to just read and understand that the logic of that program is not very difficult to fathom. But our purpose is not to learn cobalt programming our purpose is to learn information system. You agree that this kind of analysis and design is important before you undertake even writing a simple program. My humble suggestion is that in any endeavor where you have to write programs it might be useful to go through these steps formally. They will help you validate your own assumptions and so that you will make less mistakes in the program later when you write that program. We will very quickly look at the course organization. We have about 15 weeks of formal interaction. The slot for this course is six. As I explained in my notice the slot is notional. It is for avoiding examination clashes. So please ensure that you are not registered for any other course under slot six. That is important because if the exams are for this course and other course are kept at the same time then there could be a clash. That will not be possible. The formal sessions will comprise of the following topics. We will have an information systems view where we will talk about type and organization of system such as this session itself. We will then look at technology of databases, networking, web. We will also look at computer storage and I.O. technologies to understand how exactly these information pieces are handled inside. And then there will be sessions on methodology where I said software engineering, project management. So we will discuss these issues. We will also discuss implementation and operations issues of large systems. We will have some tutorials in between. I will announce the complete schedule appropriately where we will try and solve problems. All of you will be required to do course projects because this is not a hypothetical course. So you will have to practice what you are discussing and studying. So you will have to actually prepare an outline of an information system. You have to submit a detailed analysis. You have to do a design. All that you don't have to do is write programs. But since you are going to study databases I would expect you to write a square schema definition, a square queries and such things. Okay, so that's part of the course. The time required is about 40 hours of lecture tutorial sessions as is normal in any semester. Discussion sessions, self and group work would require at least about 30 hours for homework and assignments and about 45 hours of course projects. So this is the contribution that you will have to make outside the classroom. We have 5 hours of examination, 2 hours of mid-sem and 3 hours of end-sem which is mandatory. And the total work time is approximately 120 hours. I think that's reasonable. But that is the minimal that will be required. Less than that will not be sufficient. The evaluation scheme tentatively and I am open to suggestions and modifications is that assignments will constitute 20% weightage. Course projects will constitute 30% weight. Because project is something which you handle as a group and that is where maximum learning according to me happens in my experience. The project will have maximum weightage. There would be a minor project and main project will possibly combine these two defining different phases. So phase 1 could be a minor project, phase 2 could be a major project. And then of course written exams as per the senate mandate. To ensure that the practical work that you do has a larger component of evaluation the final exam will contain elements of the projects as well. So we handle it typically in two different ways. One will be there will be a question on the project that you yourself are doing. The other will be a question on some other project which is happening for which a presentation has been made here. So we will go through that. All of our exams incidentally are open notes examination. No books are permitted. But handwritten notes or printouts which are available as tutorial sheets or lecture printouts etc are possible. Let me also tell you those of you are not used to open notes exam. Open notes exams are tougher than closed notes exams because nothing of I mean describe something something etc. Such questions will not happen. There will be problems which you will have to solve. So often open notes exam generally means that you would have to read those notes very many times at home before you come here. Otherwise during the exam you are mostly reading the notes and time flies by. But still you have the access to those notes. Your own notes are permitted. No exchange of materials permitted. No carbon copies of notes are permitted. You cannot ask your friends to write using a carbon or something. But generally the idea is that you do not have to mug up any details of whatever the course discusses you should be able to solve problems. The projects will be done in groups and teams. I would expect three to four members to form a group. The group formation is entirely up to you and I would expect you to start forming these groups now so that when you come here next time you are ready with the identification of your group members. Different groups will form a team. Typically three groups will form a project team. There could be four groups also. I will decide based on the nature of the project. The project will be assigned by us once you form groups and the teams will be formed. Typically by taking your views we are basically trying to give you an exposure that if you want to understand how companies like Tarakansan Services or Infosys or such people do large projects for clients who are across the group somewhere who define the requirements and some people here do the analysis, design, coding, testing and deliver the working system. You should get a glimpse of that process. It is best if 12 to 15 people form a group and try to handle a larger activity. We will perhaps try and define a still larger project which itself is defined to be done by two or three teams each team handling a different aspect so that you understand what collaborative work is. Building information system in modern days is effectively collaborative work. So audit students will have to form a separate group. They will not have to appear for exams but they will have to do the project. The project for them will be to audit the project work done by others. So that is the way you learn without doing much work. The project activities will comprise formulation of the problem statement or system analysis producing a document which is called system requirement specification document. We will describe all of these in details later defining acceptance test criteria. How will you say that the information system is fully developed and ready and then doing a system design so you will have to produce a design document given, give a user interface usage manual. The development which is coding and testing will be limited to only writing SQL scripts and SQL queries which we shall be studying in details. Those of you who are interested in writing Java programs, HTML, etc. of course go ahead and code the whole system either now or later but that is not part of the evaluation as far as this course is concerned. These are the two references that we will use software engineering, a practitioner's approach. This is a book by Roger Pressman. This is considered the Bible of information technology professors today. This tells you very practical aspects of software engineering and database systems concepts written by Carl Silvershart and Sudarshan. Sudarshan is our own professor Sudarshan of Department of Corporate Science and I do not know whether this is the latest edition there might be a sixth edition but just cross check whatever it is. So that is the end of first session