Andreas Weigend | Social Data Revolution | Fall 2015
School of Information | University of California at Berkeley | INFO 290A-03


external image internet-OT.jpg external image the-internet-of-things-e1397666259875.jpg

IOT Farming, Appliances, and Clinics

- Robin Kalia



external image IoT-Car.jpg


Problem


At present, the various sensors embedded in our wearable smart devices, home appliances, and automobiles are generating data continuously at an enormous rate that far exceeds the user created content on social media and other platforms. Out of this massive data that is being generated continuously, more than 90% of the information is not being used, and thus, the companies do not really have a business model to utilize that information. In health care, there is abundant opportunity that needs to be exploited. A lot of practices in the healthcare industry are following the same archaic model for the past several decades.

The focus is on application of Internet-of-Things (IOT) sensors and data mining in Agricultural Farming to improve the quality and quantity of food grain, vegetables, and fruits production. At present farmers and companies have to estimate the required level of minerals required by the soil in their fields to grow specific crops and add corresponding fertilizers in an ample amount to make sure that the harvest is good. They also have to continuously spray pesticides to prevent infestation of crops. This results in crops that have unnatural levels of nitrates and other chemicals in them, and this quantity is quite harmful for living animals to ingest. Genetically modified crops are further more harmful for our bodies. Also, in a lot of cases, it is not possible to control pests since their movements and infested areas cannot be predictably tracked in-time to prevent the loss of crops on a large scale. This problem is especially predominant in South America and Indian sub-continent.

external image Precision-Agriculture.png

Hypothesis


The main target is to investigate world-leading operations performance in natural resource efficiency, environment and human health and safety.
This principal aim is to provide farmers and companies involved in agricultural farming real-time information regarding the current conditions of the soil in the fields and the speed and spatial extent of infestation of crops by bugs and other harmful creatures.

The plan for farming is to provide small embedded IOT sensors that will be placed in the fields. These small sensors use solar powered batteries and send data to servers via beacons placed alternately throughout the fields. These sensors can analyze the nitrate and phosphate contents of the soil including other factors, such as humidity and temperature of the soil and surroundings. Additionally, these sensors can record the patterns of movement on the soil that result from rain, creatures and animals walking on the soil, and movement of parasites. This data will be continuously transmitted by these sensors to servers on the backend side via the beacons. The servers will process the information regarding soil mineral composition and movement patterns of external agents and send real-time analytics report that the farmers can view from the comfort of their home on their computers or smart devices via internet browsers or a custom app. They will log in the service and they will be able to see a real-time analytics dashboard that displays the current condition and composition of minerals in the soil along with the pattern of movement of external agents in their fields.

This will help them to take timely action and make sure that they can spray the additional fertilizers only when required instead of unnecessarily saturating the field with unneeded minerals. Also, this will help them to take prompt action depending upon the pattern of movement of external agents that they can see on the dashboard to spray pesticides to deal with bugs or handle other creatures in a timely fashion that would otherwise destroy their crops and lower the corresponding yields. The main expectation for this technology of combining IOT sensors with data mining and streaming is to help farmers and companies improve their crop yields with a better quality. Thus, the aim is to improve both quality and quantity of agricultural yield, and this will have a huge impact on the general well being and health of our society. Our bodies perform according to what we feed them and we deserve to treat them in the best possible way.

external image it-system-in-healthcare-iot-online-com.jpg external image harvestgeek.png

Action


We could develop a scalable data analytics platform to provide real time data mining, seamless data streaming, pattern recognition, and recommendation system information to companies in the space of Healthcare Analytics, Internet-of-Things (IOT) data generated from sensors embedded in wearable/mobile gadgets, static/home appliances, and automobiles, Digital Advertising, and Cyber Security for increasing Business Intelligence. Using this platform, we can develop enterprise platform based products that can be used by several hospitals and small healthcare clinics and disrupt some of their archaic and mundane processes.

We could work on 2 types of products -
a) Health Care: At present, when patients are admitted in hospitals, nurses have to go to their beds and monitor their health signs every few hours. This includes signs, such as heart rate, blood pressure, sugar and temperature. Doing this task in a sequential manner for hundreds and thousands of admitted patients requires substantial manpower. Now we have smart devices with many embedded sensors that can automatically monitor these signs 24 hours a day and send that data to servers. We could work on developing an enterprise platform for hospitals where nurses can have access to a dashboard that displays all these vital signs in real-time for different patients admitted including statistics, such as mean, median, mode, standard deviation of these vital signs along with clustering and recognition of different patterns to provide a better analysis of their health data information. The problem that we are trying to solve is to reduce the high human-resource usage of overburdened hospitals and small clinics and in the process reduce the stress of nurses, doctors, and other relevant staff working there. We could sell this monthly/annual subscription based health data analytics platform to several hospitals initially that has a customized interface as per their needs.

b) Internet-Of-Things: At present, we have tens to hundreds of sensors in home appliances, such as an electric cooking stove, a washing machine, a thermostat, a smoke alarm, a refrigerator, an electric led lamp, an electric kettle, a cooking oven, a dryer, and an Air Conditioner; and in automobiles, specifically cars and trucks. The sensors embedded in these home appliances are continuously sending data to servers regarding our patterns of use and their simultaneous states via their synchronization with custom apps in our smartdevices, such as smartphones, smartwatches and tablets. Some of the major international players in this home appliances market are Samsung, LG, and General Electric. Similarly, the sensors embedded in latest cars that can sync with custom apps in smart devices, such as smartphones, smartwatches and smart tablets, send data to company servers regarding the real time state and performance of different components of their vehicles while they are being driven. This sensor generated data can give those automobile companies a lot of information regarding the performance of their vehicles. Some of the major automobile companies that have already started collecting this sensor data through custom apps are Honda, Audi, GM, Ford and Toyota. We could work on developing a realtime analytics platform to help these companies with vital information, such as the variation in performance of different components resulting from daily frequency of use of their automobiles along with the nature of interaction with them. We want to provide them quantitative information regarding the engagement efficiency of their products that could help them with a Business Intelligence platform to improve their products leading to an improved monetary efficiency.

We could target hospitals and small health care clinics for developing a monthly/annual subscription based enterprise platform whose interface is customized according to their needs. Our other target clients in the spectrum of IOT space are companies in the area of home appliances, such as Samsung, LG, and General Electric, and in the area of automobiles, such as Honda, Audi, GM, Ford, and Toyota, who could utilize this monthly/annual subscription based enterprise platform that would help them to provide data analytics information for Business Intelligence from mining the digital exhaust produced by several different sensors in their individual products.

external image sensors-connected-iot-car.jpg


external image internet-of-things.jpg external image medium_SmartHome_0103_imagestyle_ENG_1.jpg

Measure


Issues: There are a lot of companies in the field of digital advertising, e-commerce, such as Amazon and EBay, cyber security, insurance, IOT sensors, health care, banking, such as hedgefunds and traders, climate analysis, and bitcoin mining, which have access to huge amount of data from humans and their behaviors. However, most of this data is too unstructured for them to utilize it efficiently. In the case of automatic device generated data, such as that by IOT sensors, more than 90% of this data is not utilized. There is a need for companies that could help them with a better interpretation of this data that could assist them to come up with better and smart business decisions to improve customer engagement in order to prevent competition from throwing them out of business. A lot of these companies have started using these data mining and analytics technologies, but a lot of them don’t have the necessary resources to target the information they are collecting at a rapid space. Here is where we come in. We plan to start small by tackling small issues and targeting a small set of companies, but we plan to go as big as Palantir Technologies as we start generating revenues, winning more clients, and hiring more engineers, and sales and marketing people.

Opportunity: As mentioned above, we are targeting a highly unsaturated market that has not been exploited considerably. There are very few big players in this space, such as Palantir Technologies, as opposed to say highly saturated internet search engine space where we already have a prominent dominance of Google, Yahoo, Bing and a lot of other small companies that are barely surviving on digital advertising and business models modified around this source of revenue. We could try to target customers in some untapped market spaces, such as South America and South-East Asia in particular. We can also try to have our eyes set on European and North American markets as well but we want to make sure that we don't lose out in rat race, since we have several companies trying to compete for the same clients here in USA alone. At present, we could target some potential clients in the home appliances and automobiles market along with a couple of hospitals that we would like to approach, but we are worried if any of the clients would be willing to share their personal data or even give us an appointment to talk to them about some of the issues that we can help them to solve, since this data is quite sensitive and very difficult for them to trust with unknown guys. Things seem more challenging since we would be focusing on a B2B model as opposed to a B2C model where we have different modes of reaching consumers, such as kickstarter campaigns, crowdfunding, and social media, such as facebook and twitter, to spread the word. For this reason, we would need good partners to help us with formulating a strategy for reaching those clients.