Innovation
Engineering Services
- Home /
- Service / Solution /
- Innovation Engineering Services /
- Features of SOLIZE
Converting Latent Strength to Competitive Edge in a Shifting Environment
More than ever, a company’s ability to maintain a competitive edge is a test of its ability to continuously evolve and reform amid a landscape of dramatic shifts in market and other external conditions. At SOLIZE, we use our proprietary visualization and knowledge formalization technologies, alongside our capacity to implement changes to identify latent customer strengths, and convert those to competitive advantages.
The Revolutionary SOLIZE Impact on Competitive Edge
All companies must anticipate the coming era, and constantly reform and revolutionize themselves.
Rather than making an effort to achieve a transient effect, we at SOLIZE believe this revolution is an ongoing drive to create conditions for progress. We place our revolutionizing activities within a lifecycle of
SOLIZE identifies latent customer internal strengths and converts these to explicit knowledge using proprietary visualization technologies that incorporate tacit knowledge discovery and fact-based analysis. This illustrates the customer’s fully-realized form, based on genuine strengths. We also revolutionize work into a reproducible form, using implementation technologies such as 3D and IT digitization and knowledge conversion for personal and individual absorption. Furthermore, in order to increase added value and enhance competitive edge for customers, we use our strengths in implementation and adoption for conversion to organizational knowledge.
Creating a New Form of Harmony Between Humans and AI
SOLIZE splits the duties of on-site experts into “work” and “judgement”, revolutionizing these duties with IT conversions and reproducible work.
On looking intoexperts’ duties, we found that about one tenth of their work involved judgementsw, and that expert workers follow a process of making judgements based on experience. Furthermore, another four tenths of their duties consisted of “work” that followed the same logic every time, and the remaining half were “selective judgement” that even an inexperienced worker can handle if given simple options.
Moreover, we identified four stages of human behavior that flow into each other: “awareness,” “interpretation,” “decision,” and “execution”. Of these, veterans use tacit knowledge from their massive levels of experience in both “decision” and “execution.” Using proprietary AI technologies, we can now reproduce some “decisions” that were extremely challenging to parse using conventional methods.
Using these two approaches, SOLIZE will structure new duties for harmony between humans and AI.
Constructing an AI Alternative to Experts
One case of AI reproducing a expert’s work site expertise involves the automation of 3D printing quality management, analysis, and the handling of powder molding at the SOLIZE 3D printing plant.
This type of laminate molding process involves finishing work after molding, cooling, and removal. Molding takes about 10 to 20 hours. This does not mean that simply flipping the 3D printer switch produces good products, however; in rare worst cases, bad products need to be revised or re-molded. In order to produce high-quality molded products, expert engineers need to check the powder surface routinely and adjust the parameters when quality seems to be at risk of dropping.
By teaching AI this operational expertise and automating detection and parameter control for the powder surface, we are working to increase molding quality while reducing engineer and revision man-hours.
Identifying quality instructor data is a key point. Veterans can catch warnings of product quality issues from the powder’s surface condition during molding. Close observation to decode their perspective in the workplace has revealed that there are 12 patterns that powder surfaces follow, and that these can be grouped into three degrees of severity. Furthermore, we discovered that veterans can identify the cause of the issue based on its position.
Based on the categories and patterns we discovered, we created learning data from our store of molding images and implemented a deep learning model. We also repeatedly tested AI analytical results against seasoned engineer assessment results, implemented measures for means of producing outstanding characteristics via light adjustments, and constructed an AI that can substitute for a seasoned engineer’s eyes. Using this AI, we have enabled 24-hour, real-time powder surface condition monitoring.
Images alone only enable detection at the level of human recognition. However, we also found that roller vibrations and temperature were parameters with major impact on quality, and are now working on a combined detection system for these which will surpass human abilities.
Recording Tacit Knowledge and Linking it to the Future
As a technical theme for the future of AI, we hope to incorporate 3D geometry and voice data alongside present data, text, and images. When AI gains the ability to handle the 3D data of the record materials that contain dynamic knowledge, it will bring us glimpses into a new world. Furthermore, we also hope to convert tacit knowledge to explicit knowledge from voice data, which is densely packed with tacit knowledge and abundant on work sites during design reviews or otherwise.
Ten years after the problems of 2007, Japan’s industrial sector risks losing techniques due to aging veteran workers, a declining labor pool, and disinterest among the youth in manufacturing. What can SOLIZE do about this threat? How can we contribute, with our technology used to convert the tacit knowledge of expert engineers into explicit knowledge?
We will contribute by recording every scrap of tacit knowledge on the work site alongside past input, and pass this on to future generations. This is the SOLIZE social mission and a challenge we will continue to attempt as AI sector shapes our future.
TEL：+81-3-5214-1919 (Sales Division) /
E-mail：contact.si@solize.com