Do you know the running time of the dishwasher in winter night? Many people may be bored. I didn't find the answer to this question until 2013. I remember I was showing my wife excitedly with my laptop. She was puzzled. Is this kind of thing worth my excitement? The exciting reason is that Ubi (intelligent voice control device) can successfully record the sound changes in the kitchen throughout the day, I can know the sound volume and duration of the dishwasher when it is running. I can even know that it has three working cycles, during which there is a few minutes of quiet time. By analyzing the sound level of the dishwasher and reading related logs, I was able to infer the duty cycle of the dishwasher. However, what I can do with this information is another matter.
The secret life of the dishwasher
Since then, the average number of devices and sensors in our family has grown exponentially. Many networked devices are equipped with sensors. When these sensors are combined, we can understand ourselves and our lifestyle. It is for this reason that we have added microphones, light, humidity, barometric pressure, and temperature sensors to UBI devices. In the future, we hope that machine learning can keep up and provide more insights. The Internet of Things can help us better understand ourselves and our lifestyles, and take action to help us achieve our goals in life; the application of artificial intelligence to the Internet of Things can make us feel better, save electricity and maintain good health. Three objectives are needed to achieve the above goals:
1. Cost reduction and large-scale application of miniature sensors
2. Data collection and data storage cost reduction
3. Commercialization and ease of use of AI and machine learning API platforms In 2011, I was fascinated by the Twine smart device on the Kickstarter crowdfunding website. The product has built-in sensors for temperature, humidity and acceleration, and can report data via wireless network. The reason for buying it came from the fact that when I returned home, I found that there was water in the kitchen. It turned out that the refrigerator was broken. It would be nice if I received an early warning that the temperature of the refrigerator was falling! Twine smart devices can create simple rules and send emails or SMS messages when certain thresholds are exceeded. The Twine smart device was expensive in the past, and the price of similar devices has fallen sharply. GPS, WiFi, Bluetooth, acceleration sensors, infrared sensors, microphones, magnetic field detectors, force sensors, and barometers can be integrated with multiple sensors on a chip, and It is already equipped on billions of smartphones, and it is easy to implement applications through them. When collecting sensor data for Ubi, we must construct an infrastructure that can handle HTTP long polling, sensor data information streaming, data accumulation, rule-based processing, storage, and recall of data for analysis. When shrinking the graph, there may be problems with calling too many data points. Due to this problem, we have caused server crashes many times in the early days. We also need to understand that sampling data 5 times every 10 seconds may cause a large amount of data to flood our servers. Today, AWS, Google App Engine, and other companies have IoT platforms that are very easy to set up data collection rules compared to five years ago. The new requirement is to use these data to predict what we will do next or try to influence our next work. To do this, we need to gradually increase the level of information. Haeckel proposed such a level:
â—† Raw data
â—† Information
â—† Information
â—† Knowledge
◆ Wisdom In the example of a dishwasher, the original data is the sound decibel level and time; the information is to know where to collect the data; the intelligence is that a dishwasher has been turned on, is located in the same room, and can understand the operation cycle; Knowledge is the ability to determine the total length of the cycle and the moment of silence; then the wisdom is to know that the dishwasher will now run so much time and generate so much noise, so maybe I should not turn it on in the middle of the night. Collecting this kind of information today requires users to do a lot of learning and input to upgrade the system. This is where artificial intelligence can be applied, but it is a hard job to build AI for each specific scenario. Companies with limited resources need to focus on where the real opportunities are: ◆ Provide insights to users in relevant situations ◆ Combine data to create new insights ◆ Predict what will change emotions and emotions The new rules are companies that can affect users ’emotions and emotions Will win. Looking for patterns Although capturing and recording raw data has become a must-have feature for IoT devices, and being able to mark locations is another advantage to start extracting available information, companies can turn this information into intelligence in some simpler ways. In other words, they are achieved through abstraction, averaging and comparison. Abstraction can mean that we perform some kind of interpretation of information to identify events, or we integrate or differentiate to collect sums or rates.
For Ubi, it may be changes in lighting, how many times a day speaks to the device or the device speaks to the user ("interaction"), the amount of temperature change, reaching a threshold, etc. The average is also an abstraction, but can be used for individual users / devices or a larger set of users or devices. Finally, comparing the data of a specific user or device with the average value can provide a lot of actionable reference insights, all of which can be done without any machine learning or AI systems.
However, it is more powerful to allow the system to be trained to recognize and mark events. Nest has done an interesting job in visual processing. They basically implement machine vision by allowing users to draw an area on video feedback and name it as an event. For IoT device companies, if having a user tag or identifying an event will bring direct benefits to the user, why should n’t it be used to train the system to automatically recognize the event? Sound detection, whether at home, heating or air conditioning failure are all useful events that allow users to train the system. Then, the data set can be applied to tools such as TensorFlow, and another round of verification or correction is presented to the user. Particularly useful information in smart home identification includes: ◆ At home / out of home ◆ Arrival / leave time ◆ How many people are at home ◆ Sleep / wake time ◆ Meal time ◆ Use of equipment ◆ Other family activities (such as watching TV, cleaning, cooking Etc.) Putting them together one step further, we can start to combine the above information to create "knowledge" and eventually create "wisdom".
This is where machine learning can be applied to help extract prediction information. Target provides a good example of how they can predict a woman ’s pregnancy based on events. For example, you can start predicting that a family usually eats dinner at 6:45 pm. This information can be used to trigger a meal reminder at 5:30 pm. The system can also start to test input and evaluate whether the effect has a positive effect on the user. In the dining idea scenario, if the user adopts this idea, it can be regarded as a positive influence. Some other information can also be collected as an assessment of happiness index (speech analysis, early sleep time, late sleep reduction, etc.) also related to the input of the system. Perhaps it is a terrible proposition that the training system better manipulates us. However, if our goal is to improve ourselves, for example, to provide some autonomy within the constraints of the system, such as home lighting and temperature, we may achieve great success.
10-inch tablet devices have greatly surpassed netbooks in terms of entertainment, including reading, games, and audio-visual enjoyment. In other respects, the basic operation of the 10-inch tablet computer built on the touch screen ensures that the application of the tablet computer can be well realized, and its operation performance is closer to that of a smartphone.
1.In appearance, the 10-inches tablet computer looks like a large-screen mobile phone, or more like a separate LCD screen.
2.In terms of hardware configuration, the 10-inches tablet computer has all the hardware devices of a traditional computer, and has its own unique operating system, compatible with a variety of applications, and has a complete set of computer functions.
3.The 10-inches tablet computer is a miniaturized computer. Compared with traditional desktop computers, tablet computers are mobile and flexible. Compared with Laptops, tablets are smaller and more portable
4.The 10-inches tablet is a digital notebook with digital ink function. In daily use, you can use the tablet computer like an ordinary notebook, take notes anytime and anywhere, and leave your own notes in electronic texts and documents.
10 Inches Tablet Pc,Tablet Pc Android,10 Inch Quad Core Tablet,Tablet 10 Inch
Jingjiang Gisen Technology Co.,Ltd , https://www.jsgisentec.com