 Hello, my name is Professor Kavi Arya, I am faculty at the computer science department at IIT Bombay and I teach a subject called embedded systems which is basically computers which are embedded inside devices rendering the devices a bit intelligent and increasingly we find these things networked together forming systems of systems and leading to a very different kind of experience of computing than we traditionally used to have. So, what this course intends to do is that it intends to sensitize the participant to the kind of approach which is needed to understand these kind of systems and points them out to what is interesting about these systems and how to acquire the skills, what skills and how to acquire the skills to build these kind of systems and the increasing emphasis is on project based learning of these skills. So, let us go on to this. This is a slide that indicates a number of domains like the microscope shows say biology, pathology, the handcuffs show forensic applications, the tank shows military applications and so on. And a favorite question I have to students is which of these domains do not have an embedded system in it. So, typically you have students saying the cycle for instance which is down here or the scarecrow which represents agriculture. But we will have you know that agriculture is an area which is increasingly seeing the advent of robotics especially abroad. If you look at the John Deere website, you will see a lot of agricultural equipment have got embedded systems and robotics integrated into them. So, from the entire life cycle of a plant from seeding crops to harvesting to maintenance of crops, taking care of them crop care and so on right up to post harvest processing and stuff like that is handled by robots. And this is an increasingly important skill even for us because we are looking for instance at IIT Bombay on urban agriculture where we want to grow plants on rooftops of buildings in a very distributed way. And robots and embedded systems are extremely useful to bring down the cost of these kind of installations to reduce the manpower that we need to manage them because in cities manpower is very expensive and how to maximize the output of your produce. So, there has been a shift from looking at just computers which are standalone computers to embedded systems to networked systems on chip to now systems of systems. There is a complete change in device interaction. There is a growing number of critical applications and this is leading us to explore systems engineering outlook and look at multidisciplinary skilling of our students. So, the plan is in this course or at least in the short presentation to give you a flavor of embedded systems. Show you some examples, talk a little bit about microcontroller based programming, talk to you a little bit about real time support for embedded systems and two new approaches for building embedded software like model based design and new languages like Esterel and so on. I shall just mention these now in passing where these things are gone into more detail on the course, but the most important thing that we have with us is the project. Typically mid-sem and end-sem are about 40 percent, quizzes and assignments are about 20 percent and 40 percent of the course is a project and this project is very important because it is through project based learning that you actually acquire the skills. And this came about some years ago when we tried to teach embedded systems through the distance education program, it did not work too well because we realized that students need something in the hands to program and that is where we devised this, but I will speak about this a few slides down also. So, what we try to do is that we sensitize the participant to common design metrics with which we kind of understand embedded systems. Things like non-recurring engineering cost means the cost which you do right at the beginning in order to design the artifact or to conceptualize it. And after that is the notion of each piece cost which is the cost of the material and actually producing the device that leads to a unit cost where chips have a very high NRE cost almost millions of dollars while the unit cost is very low. While if you take off the shelf processors and build a system your non-recurring engineering cost is low, but there is a chance that your unit cost might be a little bit higher. Size is very important in terms of the amount of memory or the amount of gates that is the compute power that your device needs, performance, execution time, power. The more fast, the more big is your system, the more energy it requires and the less will be your battery life. So, you have to keep this in mind when devising your systems. Time to prototype has become extremely important because if you do not push your device or your prototype out in time nowadays often your design cycle times are like 6 months as opposed to a few years sometime ago you will miss the opportunity to have sales. Time to market is important. Maintainability has become extremely important. To retain, to increase the lifetime of your device in the marketplace that it should be able to be upgraded on site and correctness obviously is important and safety. The probability that the system will not cause harm. So, typically classical embedded systems were single functional like pagers and so on. They were tightly constrained as they are now also in terms of cost and size and performance and all that. And typically they were reactive and real time like the cruise controller in a car. Now, however, these things are much more complex. You have systems of systems where large numbers of autonomous systems which are independent can work together because they are networked and give you a certain experience. And so the whole characteristics of embedded systems versus SOS as we will refer to them here have changed. For instance, in embedded systems the scope was fixed in SOS it is unknown. The requirements for instance were fixed in embedded systems while now they might be changing requirements. The faults could be exceptional in a typical embedded system but now faults have to be considered as part of normal working and we have to design how faults should be dealt with as a normal case and so on. So, the new sorts of systems that we are dealing with show that many classical assumptions of systems do not really work. So, these are some of the skills that we want to sensitize the participant of this course too. And the sorts of developments that we should be prepared to address in the next few years are things like your hardware software costs are dropping. So, where you might not have considered applying an embedded system might be ripe for exploitation now. There has been a breakthrough in cognitive computing where you have newer and newer sensors which can sense things in a way that you had never anticipated before. Security has become very important and interconnection, internet of things. So, think of this intelligence in this cloud which is sensing the environment through sensors which are distributed which is then computing a response to it and affecting the process of the environment or whatever it is in some way. So, this is relevant in medical computing, manufacturing, smart homes, smart grid, smart what have you. So, there is an IDC report which says that in 2020 you shall have 20 billion intelligent devices which means almost a few embedded systems per human being on this planet. So, what are the skills that we need and how do we acquire these skills. So, some of these questions have been answered by us and we will share them with you. So, we feel that project based learning is very important where you learn by doing. So, what we have here is the e-anthro robot I will come to that in a moment. This is a robotic research platform which is shared with students who work on a laptop and learn all the skills that they need in a lab and the most important thing is a project. So, we have a project called e-anthro which is based out of IIT Bombay where we have taken the gain that we have acquired and we are sharing it with engineering colleges throughout the country where what we are trying to do is trying to trigger a robotic revolution by empowering students, teachers and encouraging entrepreneurship. And how do we do that? We do that through a number of initiatives that we have here like we have the e-anthro labs setup initiative which is targeted with setting up 500 labs in the next 3 years. We have done 104 in 2014 and we have just about 400 to do yet in the next 2 years. The robotics competition where we take a real world problem, we model it as a game with a rule book and we invite students to solve the problem using a robot that we give them. So, this is where we are unique in the sense that we have a small robot, we gift it to a selected student team of 4 students and we give them the knowledge of how to program this robot in a very distributed way through a DVD and we lead them through the process if you like of solving the problem. The solution is theirs, the process is ours where we guide them as to the various stages of solving a problem and they finally upload a video with which we choose the finalists who come to IIT Bombay. Now this has been an unparalleled success in the sense that it is growing exponentially. 2012 we had 4,500 students participate then 6,500 in the next year 12,500 and now this year 20,000 students almost. So, students seem to see a lot of value in this. Then we have an ideas competition where students can identify their own problem that they want to solve and they solve it using off the shelf components and come to the Iyantra symposium which is held at IIT Bombay where teachers come together to see what has been going on in the other labs, they get trained by us and they see these very interesting projects which have been done in the Iyantra labs and our increasing emphasis is now on sharing knowledge. So, the Iyantra resource development cell has become very important where this course will also be given to the labs who are participating in this. So, in summary what we are trying to do is we are trying to empower students and teachers and giving them the practical skills with which to solve real problems which seems to be a kind of weakness of our education system at the moment. And this is the way we do it. We have an Iyantra robot which costs about 25,000 rupees which incorporates 5 years of our experience of teaching. It has got various sensors like for instance the proximity sensors here which can sense up to 1.5 meters with millimeter accuracy. It has white line sensor at the bottom it can take various processors like Atmega 2560, Arm7 and so on and 8051 also and more important than all this is that it has got a lot of extensive software support. And I will play you a video here which sums up what all you can learn using this robot in a competition. So, this was an agricultural theme where we had urban farming where we had to design a seeding robot, a weeding robot, a fertilizing robot and a fruit plucking harvesting robot. So, in seed sowing for instance we designed a greenhouse with a number of troughs which they had to build themselves. We just give them the artwork of the flex and the basic robot which is at the bottom, but the entire system is designed by the students means this mechanism and how they have to do solve the problem. A weeding robot is also interesting in the sense that the yellow dowels are the weeds, the green dowels are the plants. So, you have to devise the mechanism using a robot where you sense and uproot only the weeds and you deposit them at a designated spot. Now, you can do it one by one or you can do it like this clever team has done build a container which they can selectively unload because they do so in a moment where it is and solve the problem very effectively and very fast. Fertilizing robot is programmed to fertilize plants in a given sequence. So, these are the plants. So, it does the black plants first as you can see and after that it will do the green plants and so on and to optimize they built two of these devices. So, they learn how to interface these sensors and actuators themselves and the next one is a harvesting robot where you have a scenario where it has to pluck ripe tomatoes while leaving behind the unripe ones. So, the entire skirt mechanism here and the gripper and everything has been designed by the students right that is the interesting thing they are only allowed to drop in this area and that area. So, he pulls the skirt in in this place and then goes and releases in that part of the the arena. So, here is how you can see students have learned very practical skills of working as a team solving a problem building a nice algorithm coming up to the fast kind of solution and they learn a lot more than just engineering in this case. So, to sum up we also teach robotics, but you learn a lot of other skills, you learn multidisciplinary thinking, you learn team working, you learn presentation skills and so on. And we are trying to evolve the students into dealing with systems of systems where they have to simultaneously optimize lots of design metrics like speed, size, power, complexity etc. Because these robots are autonomous, they have got to make sure that the battery lasts as long as they need it. They need to become Renaissance engineers with a very holistic view of the design process who should be multidisciplinary. The teams are typically one electrical engineer, one computer science guy, one guy from mechanical engineering and so on. And the skills range from things like domain specific skills to hardware, software, formal methods and what have you. And they have also to learn new system building tools, which give you a unified view of hardware and software co-design. And typically we find that design environments allow you to focus at the model based design level at a higher level of abstraction. So, typically we find that the boundary between what used to traditionally be hardware and software nowadays has blurred and where systems engineering and design thinking is a paramount skill. So, this is what we try to do on this course and we take the help of the Iyantra project with which we try to teach these concepts at a distance. So, thank you very much for listening and do look at the Iyantra website which will give you an idea of what we are trying to do on the project itself. And we hope that you will take the course when it comes out. Thank you very much. Jai hind.