Thursday, February 29, 2024
HomeAI25 AI case studies ! A summary of AI utilization cases by...

25 AI case studies ! A summary of AI utilization cases by industry [2023]

In recent years, AI has been used in various fields, transforming all industries and becoming closely related to our lives.

When considering the use of AI, it is extremely important to refer to examples. However, even if you search for examples of AI utilization, it is difficult to find out because there are too many examples.

Therefore, this time, we will introduce 25 AI utilization cases by industry in an easy-to-understand manner.


  • Examples of AI utilization by industry
    • manufacturing industry
    • Agriculture and forestry
    • fishery
    • Finance/Insurance
    • real estate business
    • Retail/Wholesale
    • Medical care and welfare
    • Construction industry
    • Food and beverage/service industry
  • Trends in AI use cases
    • Image recognition market is expanding ahead
    • The field of natural language processing tends to stay in the field of chatbots
    • The field of voice recognition is limited to applications such as generation of minutes and simultaneous interpretation
  • Examples of Familiar AI Utilization
    • Harajuku Passerby Count
    • Mercari AI Listing
    • Netflix cast selection
  • [Extra] Interesting AI Use Cases
    • Generate virtual idol face
    • Using AI in recruiting activities
    • AI that interprets sign language
    • AI announcer
  • How to refer to AI cases
  • summary

Examples of AI utilization by industry

From here, we will look at the use cases of AI by industry. The nine industries listed this time are:

  1. manufacturing industry
  2. Agriculture and forestry
  3. fishery
  4. Finance/Insurance
  5. real estate business
  6. Retail/Wholesale
  7. Medical care and welfare
  8. Construction industry
  9. Food and beverage/service industry

Two examples from each industry are presented.

manufacturing industry

AI utilization examples introduced in the manufacturing industry are ” defective product inspection ” and ” inventory optimization “.

Defective product inspection

Currently, image processing technology that distinguishes non-defective products from non-defective products is being used in the factories of various companies, mainly food and equipment manufacturers.

In addition, when linked with a robot arm, it is also possible to automatically remove defective products identified by image recognition. As a result, labor costs can be reduced, and the burdensome inspection work can be performed 24 hours a day.

For example, Kewpie Corporation, which is famous for seasonings such as mayonnaise, uses AI in its raw material inspection device for diced potatoes, which are raw materials for baby food. It has been.

Inventory optimization

A major North American manufacturer implemented AI from AI software provider, Inc., and optimized inventory counts to reduce inventory holding costs by 28-52%.

The manufacturer, which produces complex industrial products, built an algorithm that can automatically optimize the number of inventories by learning data such as the history of manufacturing orders, product configurations, bills of materials, and reorder parameters. that’s right.

As a result, the manufacturer was able to save $100-200 million annually in inventory management costs.

Agriculture and forestry

In the field of agriculture and forestry, the examples we will introduce are ” cucumber sorting ” and ” pesticide spraying with a drone .”

cucumber sorting

cucumber photo

Mr. Koike, who runs a cucumber farm in Shizuoka, created an image recognition machine that automatically classifies cucumbers according to their size and gloss.

Judging the grade of cucumbers requires skill, and before the machine was built, Ms. Koike’s mother would have spent about eight hours sorting the cucumbers during the busy season.

However, after creating the sorting machine, it became possible to sort easily, and the shipping work speeded up by about 1.4 times .

Spraying pesticides with a drone

A picture of a drone spraying pesticides in a rice field

Sakai Agricultural Machinery Co., Ltd., which sells tillage tines and agricultural machinery to local farmers, has been commercializing pesticide spraying using drones since around March 2018 and offers it as an “Agri Drone Service . 

The introduced drone is a product called “P20” provided by XAIRCRAFT JAPAN. The P20 is equipped with an AI that performs image recognition, and when it identifies a pest, it swoops down and sprays pesticides.

In addition to being able to spray pesticides without the need for human intervention, the pesticides are used only where they are needed, and as such, have many advantages, such as reducing the burden of agricultural work, improving quality, and reducing costs.


The examples of AI utilization introduced in the fishery are “ automation of feeding ” and “ prediction of catch ”.

feeding automation

Umitron, a technology venture company specializing in the aquaculture industry, has developed the UMITRON CELL, a feeding device that allows you to feed fish from your smartphone or computer.

The company also developed a fish school behavior analysis system named UMITRON FAI. UMITRON FAI is an AI that uses machine learning to evaluate fish eating conditions in real time, and is a system that can automatically determine a fish’s appetite through image analysis.

The appetite of fish changes depending on various environmental factors such as water temperature, salinity, weather conditions, wind direction, and amount of carbon dioxide. In most cases, the fisherman’s intuition is what feeds the fish, but inevitably there are leftovers and food shortages.

However, using UMITRON CELL and UMITRON FAI analyzes fish appetite data, so anyone can easily feed the appropriate amount of food.

Catch forecast

Ocean Eyes Co., Ltd. is developing “SEAoME”, which provides customized sea condition forecasts according to the needs of aquaculture farms, fixed net fishing, local governments, and fisheries research stations.

“SEAoME” is a service that provides seamless marine environment information (current situation analysis and prediction) from the coast to the open ocean using ocean numerical models, customized according to customer needs.

Changes in water temperature, salinity, current velocity, and sea level in specific sea areas such as bays can be predicted up to 5 days in advance and up to 14 days in advance.

This service can be expected to be used to prevent damage by predicting rapid tides, red tides, and sudden changes in seawater temperature, which can cause serious damage to aquaculture facilities and fixed nets.


Examples of AI utilization introduced in the finance and insurance industry are ” credit card fraud detection ” and ” stock price prediction .”

Credit card fraud detection

The introduction of AI is also progressing in the financial sector, including credit card companies.

Visa Inc. announced an analysis result that it was able to prevent damage of $ 25 billion (about 2.71 trillion yen) annually with an AI-based credit card fraud monitoring system.

Fraud detection by AI not only detects many frauds, but also has the advantage of self-learning and increasing accuracy and response capabilities.

stock price prediction

With Nikkei 225 AI prediction GROWN, you can use AI to predict the Nikkei stock price for the next month.

GROWN analyzes Nikkei Stock Average charts for more than 7,000 days over the past 30 years, and based on the results of pattern recognition, calculates the next month’s Nikkei stock price forecast, and the amount of data analysis is overwhelming.

The official website has announced that in the past four years, the probability that the stock price was expected to rise or fall, or reached the expected stock price, was “75%” or more.

real estate business

Examples of AI utilization introduced in the real estate industry are “ searching for the best housing ” and “ judging the price of land and real estate ”.

Searching for suitable housing

“AI move”, developed by Future Property Co., Ltd., is an app that uses AI to suggest the most suitable home for the person.

AI learns the user’s preferences such as the age of the building, area, and floor plan, and recommends properties that are considered to be the closest to the user’s preferences, greatly reducing the time and effort of searching for a room.

Another advantage of AI move is that you can easily complete the contract within the app.

Land and real estate price judgment

“BRaiN” operated by Real Estate Institute Co., Ltd. is an estimated land price calculation system that supports land acquisition for condominium developers.

In the price assessment when acquiring land, it is possible to solve the problems of differences in market sentiment between people in charge and unclear market prices.

When you enter the location of the land you want to buy, AI estimates the price from real estate supply data. You can create a report that includes information on the surrounding real estate and area.


The examples of AI utilization introduced in the retail and wholesale industries are ” demand forecasting ” and ” unmanned checkout stores .”

demand forecast

Conveyor-belt sushi chain Sushiro uses AI to predict demand in 1 minute and 15 minutes.

Hundreds of millions of sales data collected from each store is accumulated, and high accuracy is achieved by taking into account factors such as store congestion and customer seating times.

In the past, all analyzes were done using Excel, but with the introduction of AI, we are now able to analyze big data more flexibly, contributing to the reduction of food waste, marketing, and product development.

Unmanned cash register store

Photo of "TOUCH TO GO"

“TOUCH TO GO (TTG)” at Takanawa Gateway Station is an unmanned cash register convenience store installed in 2020.

The store is equipped with countless cameras, and AI recognizes the images projected there, so the risk of theft can be prevented.

At the time of payment, if the customer goes to the monitor, the product will be automatically scanned. All you have to do is pay and leave the store.

Although it is not completely unmanned because it still requires manpower to stock the shelves, we have been able to reduce labor costs by unmanning the cash registers.

Medical care and welfare

Examples of AI utilization introduced in the medical and welfare industry are ” image diagnosis ” and ” nursing care robots .”

Diagnostic imaging

RIKEN National Cancer Center is using AI image recognition to detect early gastric cancer.

Early gastric cancer has a variety of shapes, and it was difficult for even experts to recognize it. Therefore, using image recognition technology that utilizes deep learning, we have established a highly accurate display method with a positive predictive value of 93.4% and a negative predictive value of 83.6%.

Nursing care robot

Photo of Aiolos Robot, an AI-equipped autonomous human support robot

Aeolus Robotics in the United States has developed an AI-equipped autonomous human support robot, the Aeolus Robot .

The Aeolus robot, which has also been introduced in Japan, is used at nursing care sites to recognize residents with its object detection ability, and to monitor seizures and falls with its biosignal detection function.

Also, since it can work with Amazon Alexa and Google Assistant, you can give instructions to the robot by voice.

Construction industry

AI utilization examples introduced in the construction industry are ” automation of inspections ” and ” detection of voids in roads .”

Inspection automation

Photo of inspection of river bank protection using AI

Yachiyo Engineering Co., Ltd., a construction consulting company, has built a system that uses AI for the maintenance and management of river revetments and improves the efficiency of inspections such as damage.

Until now, the state of deterioration was inspected visually by humans, but inspection and improvement required skilled technology, which required a lot of time and money.

However, when this system is introduced, the number of man-hours required on-site is reduced to 1/5, and it is possible to perform inspections with an accuracy comparable to visual inspections by engineers.

Road cavity detection

Kawasaki Geotech Co., Ltd. has developed a system that uses an AI engine developed by Fujitsu Limited to automatically detect cavities with a high risk of road surface depression through deep learning.

With this service, by using AI to analyze the huge amount of image data collected by ground penetrating radar detection equipment, it is possible for specialist engineers to quickly and accurately discover underground cavities that cause roads to collapse.

It seems that the time to determine cavities can be shortened to 1/10 of the time of visual inspection.

Food and beverage/service industry

AI utilization examples introduced in the food and service industry are ” sushi price determination ” and ” automatic pizza making machine “.

determine the price of sushi

Yubo Co., Ltd. has opened ” beeat sushi burrito Tokyo ” , an automated restaurant specializing in sushi burrito, in Akihabara .

All prices for sushi burritos are market prices, and AI determines the price (780 yen to 1,300 yen) depending on conditions such as the ingredients of the day’s menu and the time of purchase.

In addition, there are no cashiers or clerks who take orders in the store, so you can order and receive quickly with cashless payment because you order and pay online.

automatic pizza machine

AI pizza machine

Picnic has developed an AI automatic pizza maker that can make 300 pizzas an hour.

The product measures the diameter of the pizza dough and uses an image processing algorithm to place the toppings in the optimal arrangement.

It is now possible to customize and manufacture a large amount of pizza with consistent quality, and minimize food loss due to sauces and vegetables spilling, so pizza can be produced efficiently and without waste.

Trends in AI use cases

First, I would like to introduce three trends in the use of AI technology.

  1. Image recognition market is expanding ahead
  2. The field of natural language processing tends to stay in the field of chatbots
  3. The field of voice recognition is limited to applications such as generation of minutes and simultaneous interpretation

I will explain each in detail.

Image recognition market is expanding ahead

AI technology is being used particularly in the field of image recognition . Deep learning technology, which enables image recognition, which was difficult with conventional technology, can be said to be the technology that attracted the most attention in the 2010s.

Image recognition technology is being used in cutting-edge fields such as self-driving cars, but it is also being used more and more in our everyday lives.

A major use case for image recognition is anomaly detection .

Abnormality detection can be used by learning a large amount of images of normal and abnormal conditions, and inspection work that was conventionally done visually can be streamlined and automated.

Today, anomaly detection is used not only in the manufacturing industry, but also in various other fields, such as detecting cracks in roads and bridges, and detecting suspicious persons.

In addition, the use of “AI-OCR” technology is also progressing.

AI-OCR technology is a technology that recognizes handwritten characters through a camera and converts them into character data.

In addition to attracting attention as a means of digitizing paper-based data, which has remained a paper culture and has become a major bottleneck in improving work efficiency, it is also attracting attention for its ability to quickly digitize the content of application forms and speed up administrative procedures. It has been.

While anomaly detection and AI-OCR are both areas in which issues have become apparent, it cannot be said that other conspicuous applications are making widespread progress. is.

The field of natural language processing tends to stay in the field of chatbots

One of the areas where AI is used is the area of ​​natural language processing .

Natural language is the language spoken by humans, and research and development for recognizing and understanding human speech is currently flourishing all over the world.

Chatbot technology is also being used in the field of natural language processing .

Chatbot technology is increasingly being used as a means of substituting work that traditionally required many people to handle, such as responding to a large number of inquiries.

On the other hand, in the field of natural language processing, technological development is expected to continue.

This is because current natural language processing cannot understand the context of words like humans do. Therefore, in the chatbot area, the certainty of conversation is increased by using question options and replying with appropriate answers from the database.

However, if AI can understand the meaning of words in the future, chatbots will make even greater progress.

Jobs generally referred to as white-collar workers, such as sales and clerical work, mainly use text for their work. It is expected that operational efficiency will continue to improve.

In recent years, OpenAI, a non-profit organization in the United States, has announced the general-purpose natural language processing model “GPT-3”, and its high accuracy has attracted attention.

The movement of research and development toward the generalization of natural language processing technology is gradually gaining momentum, and attention is focused on future technological developments.

The field of voice recognition is limited to applications such as generation of minutes and simultaneous interpretation

In recent years, smart speakers have attracted attention and their potential for voice recognition has been recognized.

In addition, the iPhone, which has become popular all over the world, is equipped with the assistant function “Siri” as standard, and it is now possible to perform various operations using words.

The mechanism that can be operated using voice is called ” VUI (Voice User Interface) ” and is attracting attention.

Assistants like Siri, who do not need to use their hands to operate and follow simple instructions while working, will continue to spread.

It is also closely related to the development of natural language processing technology for processing recognized speech. If it becomes possible to accurately recognize voice as text, it will be possible to realize a technology that can substitute for various tasks, just like a secretary.

In addition, in the field of speech recognition, cases such as recording meeting minutes and real-time translation are beginning to emerge.

Examples of Familiar AI Utilization

From here, we will introduce examples of AI. First, let’s look at three familiar examples of AI.

  1. Harajuku Passerby Count
  2. Mercari AI Listing
  3. Netflix cast selection

I will explain each.

Harajuku Passerby Count

Image showing how AI counts passersby in Harajuku

Counting passers-by is a classic example of image recognition.

Intelligence Design Co., Ltd. used AI to count the number of passers-by in Harajuku.

This AI is a service called “IDEA counter”. IDEA counter can count people even in crowded situations by detecting people’s heads from camera images.

Traditionally, traffic surveyors were responsible for counting passers-by. However, even in bad weather, sitting in the same place for a long time is a heavy burden, and it can be said that it is an area suitable for the utilization of AI.

In addition, this initiative was carried out for the purpose of investigating the traffic volume in the city due to the spread of the new coronavirus infection.

Mercari AI Listing

Mercari's corporate logo

Mercari’s AI listing is an example of AI being incorporated into a service to improve UX (user experience).

At Mercari, a familiar flea market app, when you take a picture of a product you want to list, AI will automatically list up candidates for brands and product names.

In the past, it was necessary to check the information of the product to be exhibited and fill it in in detail. However, Mercari predicts and fills in the product name, category, brand, etc., without the seller having to enter the product information one by one.

This makes it easier to sell.

Netflix cast selection

Netflix corporate logo

Casting using AI for the major video distribution service Netflix is ​​one of the famous and familiar examples of AI utilization.

Specifically, we let AI learn a huge amount of data, such as trend information, the relationship between cast and audience rating, and the relationship between storytelling and exit rate, and obtain references for casting that creates works suitable for users. .

In recent years, Netflix has strengthened the production of original video works, and if more data is accumulated, it will be possible to create works that are even more suitable for users.

[Extra] Interesting AI Use Cases

From here, we will introduce four interesting AI utilization examples as extras.

  1. Generate virtual idol face
  2. Using AI in recruiting activities
  3. AI that interprets sign language
  4. AI announcer

All of them are examples that are likely to be used in the future, so be sure to check them out!

Generate virtual idol face

DataGrid Co., Ltd. has developed a creative AI that automatically generates the faces of fictional idols.

In the future, people generated by creative AI may be used as models for commercial talent and EC (Internet shopping) clothes.

The advantage of using an AI talent model is that costs can be reduced and the risk of contract termination with talent due to trouble can be reduced.

The company is still developing a system that can automatically generate not only the face but also the whole body of a person , and it is expected to be used in various situations in the future.

Using AI in recruiting activities

AI is being introduced more and more in corporate recruitment activities. For example, Softbank and Amazon have introduced AI recruitment systems.

Using AI for recruitment can significantly reduce the time and effort required to decide whether to pass or fail, and it is also possible to determine the copy of the entry sheet.

In addition, there are companies where AI not only writes entry sheets but also conducts interviews.

AI that interprets sign language

Pictures of Robohon

NTT DATA Corporation and Sharp Corporation have jointly developed Japan’s first sign language interpreting application for Robophone using deep learning.

When a hearing-impaired person speaks to RoboHoN, RoboHoN recognizes and analyzes hand movements with AI, and speaks the words in Japanese.

In addition, when an able-bodied person speaks to RoBoHoN, it can recognize the utterance and display the content on an external device such as a smartphone.

AI announcer

Spectee Co., Ltd. has developed an AI virtual announcer “Yui Araki” that learns natural pronunciation, accent, and intonation close to humans and automatically reads out manuscripts.

“Araki Yui” machine-learns about 100,000 news audios actually read by announcers with the AI ​​engine “Spectee AI”.

It has been used for television, radio, and in-house broadcasting in commercial facilities, and we can expect it to play an active role in various situations in the future.

How to refer to AI cases

Currently, there are many examples of AI utilization other than those listed in this article. Another point is that various services have been created to meet emerging needs, making it possible to introduce AI at a lower cost.

When considering the use of AI, consider whether it is possible to meet your needs by introducing an AI service. It is necessary to build an AI in

When building your own AI, it is a good idea to pay attention to what data it learns and what it can do, and then consider whether it can be used in your own company.


This time, we have introduced several examples of AI by industry.

I hope you found out that AI is being used in various fields other than what we know. In addition, as AI evolves, the hurdles for using AI in business are gradually lowering.

For example, by using TensorFlow provided by Google , the threshold for creating machine learning models has been greatly lowered.

In the future, as the number of businesses using AI increases, AI will become more closely related to our lives.


Leave a reply

Please enter your comment!
Please enter your name here

Recent Posts

Most Popular

Recent Comments