ガルム List of Products and Services

Our company conducts hearings on challenges and develops AI solutions by customizing AI to meet specific needs. Leveraging our unique technology and know-how, we offer tailored development services. We provide comprehensive consulting and development, from the development of IoT devices that serve as the foundation for data collection, to AI that gathers and analyzes data, and further to the user interface that integrates and visualizes these elements. 【Our Strengths】 ■ Extensive knowledge and technical expertise ■ Quick response capability ■ Proficiency in optimizing AI *For more details, please download the PDF or feel free to contact us.

"GarmLens" is an inspection AI package for the manufacturing industry that supports all types of defects identifiable by appearance, allowing for user-specific inspection settings. It can learn with just samples of normal products, enabling implementation even with a limited number of defective product samples. Additionally, it offers a highly customizable system that can be tailored to meet specific needs. 【Features】 ■ User-specific inspection settings are possible ■ Learning can be done with normal products only ■ The scope for customization is limitless *For more details, please download the PDF or feel free to contact us.

Our company offers solutions aimed at the introduction of AI and the promotion of digital transformation (DX), starting with quality inspection and foreign object inspection utilizing image recognition technology. Based on our extensive knowledge and numerous achievements, we provide comprehensive support, including the development of AI models and hardware. We deliver systems that satisfy our customers. We can propose solutions tailored to specific objectives and challenges, including labor-saving and efficiency improvements, data aggregation and analysis, data visualization, and reduction of defective products. **Features of Quality Inspection and Foreign Object Inspection Utilizing Image Recognition Technology:** - Stable inspection results can be obtained with minimal training data. - Support for the selection and development of imaging equipment and fixtures, as well as infrastructure construction. - Further automation can be expected through integration with existing applications. - Capable of detecting various defects such as cracks, chips, dirt, and assembly errors. *For more details, please refer to the materials. Feel free to contact us with any inquiries.*

Our company proposes a "business efficiency solution utilizing RAG" that automates document tasks requiring specialized knowledge, thereby reducing the burden on personnel in administrative and legal roles. Using an AI engine powered by proprietary technology, we can process complex materials such as laws, contracts, and meeting minutes, enabling high-precision execution of tasks like inquiry responses and report creation. We contribute to labor-saving, workforce reduction, and cost reduction in industries such as manufacturing, infrastructure, and information communication, which require large volumes of document processing. 【Features】 ■ High-precision Q&A and document output supported by AI agents ■ Ability to input documents such as laws, contracts, and meeting minutes into the AI ■ Automation of tasks like application checks, summarization, and correction of typos and omissions ■ ChatBot support for existing applications and web services *For more details, please refer to the materials. Feel free to contact us with any inquiries.

We have a case of "automatic river water level detection" utilizing AI and other technologies. Based on footage from outdoor surveillance cameras, we monitor river water levels over a wide area. It can detect various situations, including when the water level rises to a predetermined height, equipment damage, or landslides. This enables a reduction in labor costs, automation, and the easy establishment of a 24-hour monitoring system, achieving greater efficiency. [Case Overview] - Monitoring river water levels over a wide area - Capable of detecting various situations such as equipment damage and landslides - Reduction in labor costs and automation - Easy establishment of a 24-hour monitoring system, achieving efficiency *For more details, please download the PDF or feel free to contact us.

Our company has a case study on "work management through posture estimation" utilizing AI and other technologies. We estimate the actions of workers from camera footage to determine how much time is spent on each task. This can be used for automatic generation of daily reports, identification of bottleneck processes, and visualization of work hours for educational support, enabling not only operational efficiency but also data analysis for that purpose. Additionally, it is possible to notify in case of mistakes in the work process, and the range of applications is extensive. It is primarily used by manufacturing-related companies. [Case Overview] - Can be utilized for automatic generation of daily reports, identification of bottleneck processes, and visualization of work hours for educational support. - Enables operational efficiency as well as data analysis for that purpose. - Notifications in case of mistakes in the work process are also possible, with a wide range of applications. *For more details, please download the PDF or feel free to contact us.

Our company has a case of "automatic recognition of parking time" utilizing AI and other technologies. Based on the footage from installed cameras, we can automatically detect vehicles staying in the parking lot and notify their status according to the parking duration. Additionally, by changing the threshold, it becomes possible not only to detect unauthorized long stays but also to calculate customer stay times. 【Features】 ■ Automatically detects vehicles staying in the parking lot and notifies their status based on parking time. ■ By changing the threshold, it becomes possible to detect unauthorized long stays as well as to calculate customer stay times. ■ While existing vehicle detection models struggled to recognize a vehicle if half of its body was hidden, our vehicle detection model can recognize a vehicle even if only 33% of it is visible. *For more details, please download the PDF or feel free to contact us.

We have a case study on "picking management in logistics using AI and other technologies." Real-time counting of items, detection of picking errors, and identification of defective products can be performed through image recognition. Inspections can be conducted without the need for barcodes or RFID, allowing for a reduction in management costs. Additionally, by using our proprietary model, it is possible to learn with minimal training data when registering new products. This can be utilized not only by logistics companies but also for pre-delivery inspections in the manufacturing industry. [Case Study Overview] - Real-time counting of items, detection of picking errors, and identification of defective products can be performed through image recognition. - Inspections can be conducted without the need for barcodes or RFID, allowing for a reduction in management costs. - It is possible to learn with minimal training data when registering new products. *For more details, please download the PDF or feel free to contact us.

We have examples of "automatic person tracking" utilizing AI and other technologies. By combining the processes of person detection and individual identification, we can continue tracking individuals across multiple cameras. We can also identify age and gender, as well as detect suspicious behavior, making it possible to apply this technology to enhance security, conduct traffic surveys, and promote sales activities in retail stores. [Overview of the case] ■ Ability to continue tracking individuals across multiple cameras ■ Possible applications include enhancing security, conducting traffic surveys, and promoting sales activities in retail stores *For more details, please download the PDF or feel free to contact us.

Our company has examples of "automated inspection in manufacturing" utilizing AI and other technologies. AI can perform real-time inspections to check for defects in finished products. In addition to checking for damages such as cracks and chips, it can address various types of defects, including dirt and assembly errors of parts, with a track record in inspecting precision components, wire harnesses, and equipment. Moreover, by applying this technology to food, it is also possible to check for discoloration and foreign object contamination, with successful implementations in food factories. [Case Overview] - We possess a unique AI model that allows inspections to be conducted using only images of normal products, without the need for learning from defective product data. This enables high accuracy even when it is difficult to collect data on defective products. - AI can prevent errors that are inevitably missed by the human eye, allowing for labor reduction and increased efficiency. *For more details, please download the PDF or feel free to contact us.