A network of sensors installed in various parts of the city (low resolution)

Edge AI sensing

Sensing solutions by creating synergy between edge and cloud

Service overview

We will accelerate the spread and expansion of sensing solutions utilizing edge AI in various business domains for a variety of partners who will be key players in realizing solutions, thereby contributing to the provision of new value to society and customers and the solving of the challenges they face.

Features

Take advantage of the characteristics of sensing devices to develop compact, user-friendly, low-power systems and solutions. Combining edge and cloud removes the following six barriers by extracting only metadata through edge-side AI processing or working with masked images to contribute new value to various industries and solve the challenges they face.

 (low resolution)

Reduced data volumes

Minimize the volumes of transmitted data by processing at the edge.

 (low resolution)

Privacy considerations

Build a privacy conscious system by utilizing edge side analysis of data and images and sending only the metadata output to the cloud.

 (low resolution)

Latency improvements

Reduce data transfer with edge side analysis and processing of data and images to prevent communication delays by eliminating round-trips of data to and from the server.

 (low resolution)

Reduced power consumption

Minimizing the amount of network data contributes to a reduction of communication costs and power consumption.

 (low resolution)

Service continuity

Edge-side processing ensures service continuity even if the cloud connection is interrupted.

 (low resolution)

Enhancement of security

Build a security-conscious system by utilizing edge-side analysis of data and images and sending only the necessary information to the cloud.

Our products and services

AITRIOS provides tools that allow our partners to focus on their own strengths and accelerate the development cycle for high value added use cases. Using AITRIOS reduces the cost of the processes and operations leading up to solution provision, thereby enabling more efficient operations.

 (low resolution)

Verified devices

Numerous devices compatible with AITRIOS services are available from SSS and our partners.

 (low resolution)

Development environment

We provide a number of SDKs and tools to help AI developers, application developers, device manufacturers, module and system integrators, and other partners to optimize the development process.

 (low resolution)

Portal

Portal is a web application for entry point to AITRIOS service provision. Manage user registration, project applications, project member registration and role settings, and marketplace app purchases and billing details.

 (low resolution)

Console

Console is a web application that provides various functions for efficiently implementing solutions spanning edge and cloud, including deployment, retraining, device management.

AITRIOS also offers plans suitable for use in application development and demonstration experiments, so please see the developer site for details.

Use cases

Edge AI sensing technology already has various uses.

Case 1: Monitoring of the product inventory in the stores (low resolution)
01 Inventory monitoring

Example scenario: Retail/logistics


Monitor the status of product inventory and other information while ensuring the privacy of customers.

Case 2: Detecting the traffic (number of people) for different sections of the store, as well as keeping track of waiting time and other attributes of queues and product layouts (low resolution)
02 Inventory retention

Example scenario: Retail


Detecting the number of people who stop in front of each shelf on the sales floor, the average amount of time they stay, and the attributes of those people helps improve product layouts and lineups. Combining this with point of sales data makes it possible to analyze the correlation between the number of people stopping, the average amount of time they stay, and sales, ensuring a deeper understanding of improvements to be made on the sales floor.

Case 3: Detecting the number of store visitors and their attributes (low resolution)
03 Attribute/people count

Example scenario: Retail


Detect the number and attributes of customers that visit the store. The data obtained on the number and attributes of customers can be used to verify measures for improving the content of flyers and campaigns or to optimize product lineups and store layouts.

Case 4: Real-time detection of the vehicles leaving/entering the parking lot (low resolution)
04 License plate recognition

Example scenario: Parking lots/logistics hubs


Detect the license plate details of vehicles entering/leaving the parking lot in real-time. Performing recognition processing, uploading the inferred metadata to the cloud, and linking with the payment system enables a smooth parking experience without requiring cars to stop at the gate.

Case 5: Enabling smart facility operation by monitoring day-to-day building/facility equipment usage (low resolution)
05 Smart buildings

Example scenario: Buildings


Detecting how buildings and offices are used while maintaining privacy and confidentiality enables efficient facility operations, while connecting that data to systems that manage air conditioning and other equipment contributes to reduced power consumption and CO2 emissions.
Moreover, automation of visual reading of analog meters within the building through image analysis leads to labor savings.

Case 6: Measuring and detecting the most efficient time required per task, without the need of manually measuring tasks with stopwatch devices  (low resolution)
06 Detailed analysis of employees

Example scenario: Factories


Analyze the tasks within the process rather than having analysis staff check or perform measurements with a stopwatch. Edge AI analyzes videos for visualization of the time required to assemble individual parts, identification of the amount of time that workers stop working for some reason, and more.
This can be applied to such efforts as standardization of process working time and analysis of problems, thereby improving productivity and yield.

Case 7: Overall equipment effectiveness (OEE) measurement at the production facility without the requirement of customizing existing equipment (low resolution)
07 Measurement of overall equipment effectiveness

Example scenario: Factories


Measure the overall equipment effectiveness (OEE) without customizing existing equipment. Cameras are installed for production equipment input and output locations, and edge AI analyzes the operating rate and performance.
Visualizing the production efficiency of conventional equipment allows review of the production line and quantitative evaluation of the improvement effect to enhance productivity.

Next step

We will provide an example of the overall flow from order to launch of use by a partner company.

Icons to be emailed to you (low resolution)

01. Fill out the requirements and contact one of our representatives.

Icons of during meeting (low resolution)

02. We will schedule a meeting and discuss the use case you have in mind.

Icons of proposing (low resolution)

03. This will allow us to provide specific information for implementation, including partner referral.

Inquiries

For more information about edge AI sensing, please contact us.

Click here for inquiries >

*AITRIOS and AITRIOS logos are the registered trademarks or trademarks of Sony Group Corporation or its affiliated companies.