Data Science & Artificial Intelligence for ALL

Editor's Notes

• #2: Hi,
• #3: According to a report of PWC, By 2030, AI will be15.7 trillion US dollar industry; thereby surpassing the complete budget of India and China. U.S. and China are two majors in the world of AI. We are 130 crore Indians. You will be surprised to know that, 31% of Indians have smart phones; worlds average is 53%. Three years back we were 155th at mobile data consumption. Presently we are world’s number one. Just imagine what kind and amount of data we are contributing to?
• #4: The number of mobile devices will sharply increase in the near future, a phenomenon which many have termed the 'Mobile Big Bang'. It has been predicted that by 2020 there will be approximately one trillion devices in operation, and a new era, the 'Machine to Machine' era, will open up life-enhancing technologies even more by connecting mobile phones and other devices to sensors and information appliances. The amount of data transmitted and shared by all these systems will also explode. In 2009, there were only about 0.8 ZB (1 Zettabyte = 1 trillion Gigabytes) of information; by 2020, with vastly increasing volumes of digital information generated by devices such as sensors, mobile devices, business IT systems, and environmental and biomedical systems, this will increase to 35 ZB.
• #5: On 25th Aug. 2017, under the leadership of Prof. Kamakoti of IIT, Madras; an eighteen member AI Task Force was formulated to leverage AI for the economic benefit of India. The AI Task force has identified 10 specific domains that need attention with respect to AI. They are, manufacturing, fintech, healthcare, agriculture, education, retail and customer engagement, public utility services, aid for differently-abled persons and accessibility technology, environment and national security. . India, in total, has over 750 startups using AI, machine learning, big data. The current AI industry is estimated to be worth Rs. 1260 crore of revenue, annually Meanwhile in Jan 2018 the Prime Minister of India, in World Economic Forum at Davos, said, “Today, data is a real wealth and it is being said that whoever acquires and controls the data will have hegemony in the future. The global flow of data is creating big opportunities as well as challenges”.   In March 2018, The Artificial Intelligence (AI) task force suggested creating a National Artificial Intelligence Mission (N-AIM) that will serve as a nodal agency for coordinating AI related activities in the country.   The core activities include funding establishment of a network among Academia, services industry, product industry, startups and Government ministries, besides helping studies to identify concrete projects in each domain of focus. Side by side, in Jan. 2018 the AICTE new syllabus was launched. It was suggested, that, all universities and engineering colleges of India to adopt the same curriculum. Prof. Anupam Basu and his team meticulously worked and identified a thread of Data Science and Machine Intelligence in Computer Science and Engineering syllabi. This thread suggests Artificial Intelligence, Machine Learning, Data Mining, Soft Computing, Speech and Natural Language processing, Data Analytics, Information Retrieval, Neural networks and Deep Learning, Multi Agent Intelligent Systems and Data Analytics as subjects to be taught to the students from 3rd semester onwards, thereby making them experts in this field by the time they graduate. In May 2018, it was reported, that, Niti Aayog will soon come out with a national policy on artificial intelligence, outlining the scope of research and for the adoption and commercialisation of the technology While we are all gearing up to Propagate Date Science and Artificial Intelligence (DSAI) to all computer literate professionals and students of professional institutions through learning, innovation and Research & Development,
• #6: So far I have given you India’s status and World view. In the last webinars different speakers have covered the use of DSAI in the area of Business & Management, Games & Medicine. I would be telling you as on what DSAI can do and its use in the core fields of engineering. Also, just to change the taste, I have added some bizarre areas in which AI is being used. Once you are properly motived, there comes the recommendations on where to start.
• #7: Take the example of Stocks Trading & Explain Let us say shares belong to banks, white goods, hardware, software, FMCG etc. Next is finding the outlier. After 15 days for what will be the value of my portfolio? May be a new category appears, like AI related startups. What to do next? Some Shares are going down. Should I sell it/ buy it more/ hold it?
• #8: In the design of Construction Management System, the various people involved in the process are given access as per their access levels. The Management here involves managing the resources like raw-materials, inventory, crew (staff), budgeting, task management, scheduling etc. In big data analytics, the large amounts of data gathered from various resources will be stored, mined and the result will be use optimally and the resources managed tactfully. Estimation of the requirement resources is a major task for any construction company. The success of the project lies in the correct estimate of the required resources. Any civil engineering company will go for tenders to get the project. In the tender process, the main job for the key persons is to estimate the resources required like raw- materials, staff, workers, vehicles and the duration of the project. The success in the tender lies purely in the proper estimation. Without estimation, a vague or rough estimate may result in either loss in tenders or generate loss in the execution of the project. Therefore, the estimation is very crucial for the construction company. The proper analysis with the building data and the study of the building and geographical area can help us in the prediction of earth quakes before its occurrence. This prediction is possible with the big data analytics by storing the related variety of information and processing that huge amounts of information. Further, other techniques like Neural Networks, Data Mining will also help in the proper analysis and accurate prediction of Earthquakes. Hope with DSAI in hand, we will have no bridges collapsing and no lives lost; because before we create the design … we will have all possible ways in which a bridge can collapse. Also, if there is any study coming up in by the time the construction finishes or any more damage is there in any other bridge, that will strengthen the design adding one more possible way the bridge can collapse and the design will be more robust accordingly. In US, the imagery of a crack on the flyover is accessed by civil engineers and the repairs teams reach straight to the site. Even there are drones taking care of people who might drown, or boats being capsized. There are a water scarcity in Callifornia some time back and the govt. used drones to find out who is sprinkling more water in gardens than allowed. We have so many rivers to be cleaned, ground water quality deteriorating by day and sewers chocking and lives lost every day. We need to have a network of water bodies and drainage system , the way we have google maps for roads. Then we will be able to identify which water body and drainage system to be cleaned first to ensure clean water and sewers. This will be part of Indias RiverGo program which will reduce the travel time for trucks from Punjab to Vishakhapattam from 27 days to only 4 days. The traffic moves like a water flow, If we block a road, the traffic finds its alternate way. Seeing the pattern of peoples’ choices; the collected data can be trained in a way that we do not have to spend our lives in traffic jams every day hence reducing the waste to manhours and fuel. We have seen aero planes hovering as there are no airstrips and hangers vacant to land them. Better signaling will be possible, if we have proper data utilisation by storing, retrieving, analysing, predicting consequences. Imagine our railway traffic system without delay. All trains moving in time… imagine the amount of man hours saved. Imagine the convenience. Yes! It is possible. Next scenario can be With so much crowd around, there are always chances of stampede lurching around. Now, the best part is… most of them have smart phones. Identifying usual pattern of pedestrian movement, possible nearest exit locations and guidance to all individuals separately can lead to saving lives. It started with only six satellites. Right now earth is covered by numerous satellites. Humongous amount of spatial data has been stored by Geo Earth mapping companies, waiting to be explored. Be it areas where humans can reach, or difficult terrains, or emergency services or watching borders or unlawful cutting of jungles canopies or mountain or jungle fire. You name it and it is there. In Malaysia, all buildings have been digitised. They can reconstruct a devastated building in a span of few months. Amazing!!!
• #9: With the adoption of smart-connected products, the internet of things and big data are quickly becoming integral to product development. And mechanical engineers should be aware of how data can now reform designs. The internet of things (IoT) is going to change the scale of data that modern product development and manufacturing companies are creating and consuming. A tremendous amount of data will be collected from connected devices and this can be transformed into consumable information assets.” With IoT, mechanical engineers have the opportunity for product insights that were never possible before,” . “With an IoT-enabled product or device, products can stream usage data back to engineers. By applying analytics to this big data, engineers can see how products are used in the field, where customers tend to have problems, and what functions are not used,” Mechanical engineering deals with designing products that are used in the harshest conditions imaginable on the planet in minute design details to ensure the most robust end product. Before designing and building parts and assemblies, engineers need to be very thorough in their calculations and modeling. For example, factors like high temperature, high pressure, stress, vibration, corrosion, aging all act in parallel on the mechanical systems. We can look deeply into that data to better understand the combinations of these variables that are causing mechanical failures and then we can bring together the data streams for both physics and engineering. This non-linear root cause analysis shows us the real world we deal with on a daily basis. It is ideally suited to leveraging big data and analytics and it benefits multiple groups within our company including engineering, manufacturing, sustaining and maintenance. A wide variety of projects from downhole tool reliability to recruiting optimization are there, and we’ve been able to leverage a number of analytics tools in that effort including Spark, Hive, YARN, and other components For a data scientist, building a Spark-based end-to-end solution is quite fascinating, since Spark has taken some of the burden off of the data scientist that would ordinarily require more extensive software engineering skills. Data and analytics allowed us to perform deep exploration of different trends and variables which would have been impossible to correlate without advanced machine learning and analytics. We arrived at a conclusion that surprised us and could only have come from data analytics. You know, AIRSIN is the real world simulation created so that we test the new products like driverless cars in virtual world without damaging it
• #10: Today dependency of energy generation is shifting from conventional to non-conventional sources. Renewable energy technologies (RETs) like hydropower, biomass, wind and solar energy are commonly used for power generation. Direct integration of these distributed source of energy to the main grid is very challenging as it need efficient energy management strategy. By DSAI we can do the optimization during peak time, blackout, brownout and catastrophe. Also renewable energy resources (RES) are intermittent and amount varies with large swing.. Monitoring of consumption pattern is very important when we have to formulate flexible tariff plan. The consumption varies on season, days and day temperature. Like in rainy season consumption is very less as compared with summer( when most of us use AC and refrigerator ) and winter ( room heating appliances). By days we mean in the morning power is used maximum as in most of the households water pumps, mixer-grinder, geyser and other home appliances are used. So in the morning power consumption is high. By the day means some of the days are hotter and some are cooler and few of them have high humidity. On that day we use more home appliances. Which results in the increase of the power consumption. Analysis of such data is win-win situation for user and distribution companies. As it helps consumer in shifting their consumption pattern to the point when the tariff rate is least and for the distribution company it helps them to make policies and tariff breakout. Whenever there is distributed power generation and it’s integration; maintaining power quality is very important. Keeping a check on the power quality not only improves the revenue collection by controlling theft but also improves the life of home appliances and distribution system. When we say power quality, by this we mean maintaining the power factor, voltage and current. By regulating this we can check any fault present in the system. Good quality means utmost consumer satisfaction and least losses. when we talk about the smart grid then cyber security is the prime thing of concern. Everything in smart grid is digital and automated. As we know that consumption pattern of consumer is always monitored. If anyone hacks data of any of the consumer then he might be looted or theft can be done easily when no one is at his/her home. Also by hacking, anyone can transfer his/ her energy bill to someone else’s bill and will enjoy free electricity. load balancing means efficiently distributing the load or power. In this we shift the load according to the need and priorities. As we know that all consumers are producers also, which we called as prosumers, and present in rural and urban areas. In the evening, load can be intermittently shifted from rural area to the urban areas. This can also be done on harsh climatic conditions as rural people use very less appliances hence there needs are also a few. If any area is suffering from blackout or brownout then in that area power can be supplied from any of the nearby grid using optimization technique and Neural network technique. 6. As we know that in smart grid we continuously keep check on power quality, load distribution, billing as tariff is flexible, blackout, brownout and supply in catastrophic prone area. For regular monitoring and constant supply of power, data analysis is must. By doing this we can ensure that whether the affected area is again established or not.
• #11: Learning Organic Chemistry- Predicting outcomes of complex organic chemistry reactions using neural sequence to sequence models. Converting video to text- MIT-IBM labs have created 1 million videos two seconds each for gesture identification; thereby converting video to text for visually impaired population. Creating Music- Different musicians play differently, and humans can identify which one is there. Once we have the notes with us. Creating a different rhythm is what AI can do for us. Creating Movie trailer- Out of the whole movie, AI identifies the scenes which create most sentiments out of Sentimental Analysis. These scenes are stitched together to make most characteristic trailer. Identify the artwork, style or artist- Whether it was Pablo Picasso. ... Vincent van Gogh. ... Leonardo da Vinci. ... Or Michelangelo. Informed humans can identify and now AI also. & Product Authentication- You never know the source of farm produce which comes to your home. Here, Data Science, AI along with block chain can solve your problem. Identifying whether your Dehradooni rice came from Dehradoon, Kohlapuri slippers are from kohlapur or you expensive Turkish carpet belongs to turkey. Also, the 18 year old whisky from Scotland is actually 18 years old and straight from Scotland. And.. Last but not the least… whether that 11 carat diamond gifted to your by that multimillionaire is worth a fortune or fake. Smile go ahead and learn… as Learning is the easiest thing that keeps you happy!
• #12: You know, there is talking camera, and on its first day of lunch 3,000 blind people posted their photos on facebook. The speech to text and transcribed text as we do between Hindi and English and then converting and translating this to any other language is something accessibility ca do for us. The Alphago zero could break its own record by ‘Lookahead’ and ‘intuition’. One day one it was taught various moves, in three days it efficiency was less than 30%. In 30 days, where it kept playing with itself and the program kept getting robust, it was in best 30 players of the world and in 40 days it beat all of them
• #13: DSAI is for everyone. If you are not creating it you are using it and you are contributing to strengthen it. So, no one is spared, if you learn it you know the ins and outs. You know the capabilities and treats. You are better informed. Follow the DSAI thread in Classroom or as open elective. Go for various online or offline courses. Instead of patchwork learning python somewhere, R somewhere , SAS somewhere or Octave somewhere else, try getting a complete package. UAE is the first country to have a ministry of Artificial Intelligence. Let us strategically learn AI to have skilled personal by this year end and creative AI personal by next year end, so as to populate them to work in India and abroad in the interest of mankind. Let India change its perspective from snake charmers nation to DSAI workhorse.
• #14: Add hyperlinks