Workflow

Think Efficiently and Systematically

We help individuals think more efficiently and systematically to improve clarity, decision-making, and problem-solving.

Think Efficiently and Systematically

We help individuals think more efficiently and systematically to improve clarity, decision-making, and problem-solving. In this context, the analysis was conducted through workflow analysis with a tool that maps AI's internal feature processing into workflows.

By structuring activities according to levels of data processing, this approach provides a clear framework for understanding how raw data evolves into analytical outputs and ultimately into strategic decisions, while also enabling the identification of inconsistencies and improvement opportunities.

Approximately 70% of the items show alignment, indicating that the overall structure reflects a progression based on levels of data processing. However, two recurring issues are observed: idea/shop-floor activities assigned to excessively high levels and strategic/judgment activities assigned to excessively low levels. Taken together, this suggests that the L-level framework is fundamentally designed around data processing maturity, but there are partial inconsistencies where the application of the criteria is not fully aligned.

The first group (high level) shows a relatively clear structure. Activities such as problem definition, work improvement, and quality diagnosis are concentrated in L1-L2, while some analytical and planning activities extend to L3 and above. This reflects a typical pattern of shop-floor-centered to partial analysis, where raw data-based activities dominate and are extended to processed data only when necessary.

The second group (middle level) presents a mixed structure combining shop-floor focus and expanded levels. While many items remain in L1-L2, there is also a significant presence of broader ranges such as L2-L10, L3-L10, and L1-L10. This indicates that during the execution phase, activities continue to rely on shop-floor data, but also require higher-level data due to factors such as training, scaling, and system integration. As a result, this group can be interpreted as a shop-floor-based and expansion-driven hybrid structure.

The third group (low level) exhibits a diffusion pattern dominated by L1-L10. Although some items fall within L3-L10, L4-L10, and L7-L10, most are assigned across the full range, weakening the distinction between data processing levels. This reflects the nature of validation, performance analysis, and scaling decisions, where multiple levels of data are used simultaneously. However, it also reduces the clarity of the classification framework, making the L-level distinctions less meaningful in this group.

The details for each level are as follows.

High Level (Idea-Centered)

  • 658.401.001 - Defining core problems and opportunity ideas for AI-based manufacturing development
  • 658.401.002 - Selecting manufacturing competitiveness ideas instead of trend-driven AI adoption
  • 658.403.001 - Setting AI use ideas to reduce production time, cost, defect rate, and repetitive work
  • 658.403.002 - Identifying AI application ideas for production management, quality inspection, equipment management, and process analysis
  • 658.401.003 - Connecting AI manufacturing innovation ideas with short-term and long-term business strategy
  • 658.300.001 - Designing field role ideas that combine AI automation and worker decision support
  • 658.404.001 - Planning technology, budget, workforce, and data scope needed to realize manufacturing AI ideas
  • 658.401.004 - Setting expected productivity and quality improvement effects for each AI application idea
  • 658.403.003 - Collecting AI ideas from production, quality, equipment, and IT departments
  • 658.401.005 - Writing AI development goal ideas in language understandable to manufacturing sites
  • 658.500.001 - Deriving manufacturing work improvement ideas based on repetition, complexity, and processing time
  • 658.403.004 - Identifying AI automation ideas for simple repetitive production, inspection, and recording tasks
  • 658.562.001 - Diagnosing process improvement ideas to reduce defect rates and quality variation
  • 658.500.002 - Exploring AI application ideas to resolve bottlenecks in production lines
  • 658.403.005 - Designing work ideas that separate AI-applicable processes from expert judgment processes
  • 658.404.002 - Prioritizing high-impact and feasible manufacturing AI project ideas
  • 658.406.001 - Selecting initial AI project ideas that can quickly build trust on the shop floor
  • 005.800.001 - Reviewing manufacturing AI project ideas with security and legal risks in mind
  • 658.404.003 - Evaluating AI project execution ideas based on cost, timeline, data, and operating workforce
  • 658.403.006 - Finalizing manufacturing AI project ideas agreed upon by production, quality, IT, and management

Middle Level (Result-Centered)

  • 005.740.001 - Securing production, quality, equipment, and logistics data to verify manufacturing AI results
  • 006.310.001 - Confirming AI prediction and analysis results based on data quantity and quality
  • 005.740.002 - Improving analysis result reliability through cleansing duplicate, missing, and incorrect manufacturing data
  • 005.800.002 - Securing safe AI utilization results through access control for confidential manufacturing data
  • 658.403.007 - Confirming manufacturing data utilization results through MES, ERP, PLC, and IoT integration
  • 658.404.004 - Deriving AI applicability verification results through a specific production line pilot
  • 658.404.005 - Confirming pilot operation results for selected processes, equipment, and product groups
  • 658.401.006 - Measuring improvement results in production speed, defect rate, equipment utilization, and delivery performance
  • 658.403.008 - Collecting shop-floor user evaluation results on usability, accuracy, and process fit
  • 658.155.001 - Confirming reductions in production disruption and quality risk through human review procedures
  • 006.310.002 - Analyzing AI performance improvement results from adjustments to data, models, and process conditions
  • 658.403.009 - Summarizing manufacturing AI usability improvement results based on shop-floor feedback
  • 658.401.007 - Recording pilot result findings, success conditions, and required improvements
  • 658.404.006 - Calculating realistic results for operating cost, field workforce, and training level
  • 658.403.010 - Deciding whether to expand AI adoption based on pilot verification results
  • 658.401.008 - Deriving manufacturing AI performance measurement results based on predefined evaluation criteria
  • 658.401.009 - Analyzing performance results centered on production time, cost reduction, defect rate, and equipment utilization
  • 658.401.010 - Comparing before-and-after results in productivity and quality after AI adoption
  • 658.403.011 - Evaluating AI results on field workload reduction and production decision quality
  • 658.403.012 - Summarizing performance-based judgment results for expanding, improving, or discontinuing manufacturing AI

Low Level (Execution-Centered)

  • 658.403.013 - Establishing execution standards for safe and consistent AI use in manufacturing sites
  • 005.800.003 - Creating execution rules that separate manufacturing data allowed for AI input from prohibited information
  • 658.403.014 - Applying execution standards that distinguish human-reviewed work from AI-automated work
  • 658.155.002 - Executing response procedures for production and quality problems caused by AI errors
  • 005.800.004 - Executing manufacturing AI governance covering security, technology leakage prevention, privacy, and ethics
  • 658.312.001 - Executing AI capability training for manufacturing site employees
  • 658.312.002 - Executing practical manufacturing AI training based on production, quality, and equipment cases
  • 658.312.003 - Executing training on reviewing and improving AI analysis and prediction results
  • 658.312.004 - Executing customized AI utilization training for production, quality, equipment, and logistics departments
  • 658.312.005 - Operating continuous training and support channels to reduce shop-floor anxiety
  • 658.406.002 - Executing AI expansion by production line, factory, and business site based on pilot results
  • 658.403.015 - Executing AI integration with existing manufacturing systems and process operation flows
  • 658.406.003 - Executing AI application methods adapted to process, product, and factory characteristics
  • 658.404.007 - Executing expansion of cost, security, performance, and operating workforce as AI usage increases
  • 658.406.004 - Executing manufacturing AI expansion that prioritizes production stability and quality maintenance
  • 658.500.003 - Executing a manufacturing AI operation management system after initial implementation
  • 658.406.005 - Executing adjustments to AI models and usage methods according to process, equipment, and demand changes
  • 658.401.011 - Executing continuous improvement based on shop-floor feedback and production performance data
  • 658.562.002 - Executing manufacturing AI quality improvement by accumulating defect, error, and equipment anomaly cases
  • 658.401.012 - Establishing AI as a long-term execution foundation for strengthening manufacturing competitiveness