Home
Domain Specific AI powered by Takane
Takane, a generative AI for enterprise use, and domain-specific AI automate tasks that previously required manual labor. We create new business opportunities and new value through AI solutions specialized for various industries, such as manufacturing and healthcare.
Challenges in utilizing AI in enterprise
In an enterprise environment, creating an AI model for individual tasks requires specialized knowledge and takes time. Issues such as an inability to handle specialized tasks, delayed response due to increased calculation volume, and increased costs have prevented generative AI from being fully utilized in business settings.
Domain Specific AI powered by Takane
Fujitsu's Takane-specialized technology makes it possible to utilize generative AI for complex, specialized tasks. Fujitsu's strength lies in its ability to combine "Takane," the LLM boasting the highest accuracy in Japanese language translation, with domain-specific AI that condenses operational know-how from various industries, such as manufacturing and healthcare. We optimize "Takane" to meet the individual needs of our customers, and with new generative AI tailored to corporate requirements, we can address use cases and performance that have not previously been addressed.
The benefits of Domain Specific AI powered by Takane
- Manufacturing
- Automating inspection processes improves the efficiency of the entire manufacturing process and the quality of the products themselves
- Healthcare
- Freeing up patient time by reducing information seeking and medical summary creation.
- Improved safety and satisfaction by reducing the workload of electronic medical records. Standardized discharge summary quality by utilizing LLM drafts.
|
|
|
|
|
|
Demo
App
Click here to try the demo app
Technical overview
Target Industry/Users
- Manufacturing: Control design engineers in the manufacturing industry, manufacturers designing and producing complex products
- Healthcare: Medical institutions, medical professionals
Challenges in Target Industry and Operations
- Manufacturing:
- There are a wide variety of requirements that must be confirmed during manufacturing, which means that checking designed products is costly and time-consuming.
- Technology is becoming more complex, and the number of engineers with extensive knowledge is decreasing, making technology transfer more difficult.
- Healthcare:
- Medical information such as electronic medical records is vast, and much of it is unstructured data (free text, reports, etc.), making it difficult to extract and utilize the information.
- It is difficult to understand and analyze patient information over time, which is hindering the realization of personalized medicine.
Technical Challenges
- Manufacturing:
- Understanding unstructured requirements documents including diagrams, identifying the shape and location of parts without part numbers, and automating inspection processes.
- Healthcare:
- Technology for accurately extracting, structuring, and utilizing necessary information from various unstructured data in electronic medical records (progress notes, various reports, etc.).
- Interpreting information in medical records taking into account dates and time axes, and associating data chronologically.
Solutions
- Manufacturing:
- Mechanisms for correctly reading structures including tables, understanding industry- and company-specific descriptions, mechanisms for identifying component types from their appearance, and technology for generating inspection programs.
- Healthcare:
- Automatically extracting patient information such as severity, disease classification, and vital signs from unstructured medical information such as electronic medical records to build a high-quality structured database.
- Constructing a patient-centric dynamic knowledge graph (Dynamic KG) to represent a patient's clinical information over time and its uncertainty.
Fujitsu's Technological Advantage
- Manufacturing:
- Automated programming technology for industry-specific testing tasks through industry-specific understanding of requirements and the analysis and generalization of various testing know-how.
- Healthcare:
- Our proprietary KG-enhanced RAG technology deeply understands the terminology and nuances specific to the medical field, enabling highly accurate information extraction and structuring. It also captures temporal changes and uncertainties in patient condition.
The benefits of Domain Specific AI powered by Takane (Detailed version)
- Manufacturing:
- Automating testing operations improves the efficiency of the entire manufacturing process and the quality of the products themselves.
- Healthcare:
- Takane accurately extracts and structures necessary information from electronic medical records, significantly reducing the time doctors and researchers spend searching for information. This improves the quality of diagnosis and treatment, frees up time for patient care, and accelerates research.
- Improved safety and satisfaction due to reduced electronic medical record workload. Standardized quality of discharge summaries.
Use Cases
- Manufacturing:
- When designing products using CAD, checking in advance to ensure they meet manufacturing requirements reduces rework.
- Healthcare:
- Creating discharge summaries: Automatically summarizing the entire admission and discharge process reduces creation time and improves the safety of care transitions.
- Understanding patient condition, treatment plans, and examinations: Quickly understanding past medical history and test progress improves the efficiency of initial decision-making.
Case studies
- Manufacturing:
- Demonstration of automated inspection procedures for production technology requirements at an automobile manufacturer
- Healthcare:
- AI-based medical support using electronic medical records, providing information that takes into account the chronological changes of re-admitted patients
Technical Trial
Related Information
- Press releases
- TechBlog
Demo
App
Click here to try the demo app