SpectAWork innovation through the integration of tacit knowledge and natural language processing AI

Features

• 

  AI solution specializing in manufacturing

  Text searches, Q&As, and other generic natural language processing AIs are not sufficient for solving challenges in manufacturing, which requires decisions with complex tradeoffs under multiple conditions. It requires the construction of an advanced knowledge base that enhances an AI inference engine dedicated to design work such as quoting, design, manufacturing, and maintenance.
  SpectA is equipped with a specialized knowledge base for the manufacturing industry built using the manufacturing technology cultivated over 30 years at our own company. Combined with the inference engine that SOLIZE has developed which reproduces the key insights from skilled workers, it provides information detection and semantic interpretation accuracy that helps directly with problem-solving from the moment of deployment.

• 

  Achieve reliable results in a short period of time

  In sophisticated discussions involving tacit knowledge built over time, the AI’s information accuracy tends to suffer from a lack of quality and quantity in machine learning training data, and a lack of clarity about the chain of events that interest skilled workers and the logic behind their thinking (i.e., feature value).
  SpectA can be optimally tuned to the challenge at hand by externalizing tacit knowledge and data science to identify the key ideas behind skilled workers’ thinking honed over years and building a training dataset focused on solving the challenge. This allows for swift improvements in AI information accuracy based on a limited set of data and results.

• 

  Maximize the quality and quantity of human decisions

  In manufacturing, product issues can have significant downstream effects. The “basis of decisions made by AI” and “the extent to which humans can be held accountable” are two important concerns with many customers worried about AI turning into a black box, the impact it may have on weakening the transfer of skills and technologies and the division of roles.
  SpectA transforms work processes to ensure optimal collaboration between human decision-making, deep learning, and rule-based systems while separating decisions from tasks. This approach maximizes the automation rate while preventing core decisions from becoming a black box. It enhances human capabilities to improve both the quality and quantity of decisions, providing a wide range of outcomes including streamlining, creation of added value, transfer of skills and technologies to the next generation, and development of human resources.

• 

  Sustainable evolution of knowledge

  Once an AI that can be used in work is built, it needs to continuously learn and be tuned in time with changes in the business environment.
  SpectA is equipped with multiple functions for this, including those that allow easy tuning without coding, making minute adjustments manually over a wide range, in addition to being fitted with technologies that help it evolve while augmenting the inference engine and knowledge base using feedback from users on detected information. This allows the use of information and knowledge acquisition that continuously evolves to resolve business challenges.

SOLIZE’s strengths

By combining “multi-modal AI technology focused on natural language processing” with “our unique work process innovation methodology that visualizes decision-making” created based on our manufacturing technology fostered over 30 years in our design, analysis, prototyping, and manufacturing work on the field and improvement and reform technology, an environment is created where the knowhow that drives the company’s competitiveness as represented by tacit knowledge of skilled engineers is made explicit and used as organizational knowledge that everyone can reproduce. We have successfully improved QCD and shifted resources to creative work as customer’s ability to implement innovation at over 250 companies across industry borders.