Practical guides for manufacturing AI
Questions and checklists for data readiness, pilot design, metrics and ontology planning.
49 / 49
Product brief
Product roles and control flow, printable for internal review
ExploreOur vision and people
Meet Ingkle: our vision, values, milestones, people, and offices in Cheonan and Suwon.
ExploreExperience across manufacturing environments.
Manufacturing data foundations across semiconductor materials, automotive, steel and varied equipment environments.
ExploreResearch for the next operational challenge.
Develop and demonstrate data and AI technologies through industrial and SME programs.
ExploreConnect NAZARE and NAVIS to your existing systems.
NAZARE connects equipment, MES, ERP and QMS data. NAVIS uses the same evidence for analysis and response. Explore how the products work with existing systems.
ExploreOperate NAZARE and NAVIS within each site’s permissions.
NAZARE separates data access by site and user. NAVIS connects the sources, tools and approvers required for each role. Explore separated operation and execution permissions across sites.
ExploreUnderstand the operation. Act with approval.
NAVIS is an Agentic AI platform connecting users with manufacturing operations. It reasons with live equipment data, operating records and knowledge structured in NAZARE, then requests execution via MCP after manager approval.
ExploreMove from finding information to the next task.
Connect data access, business context and model inference as tools for agent workflows.
ExploreReview the evidence. Approve the next action.
Review and approve the target and evidence in NAVIS. NAVIS requests execution through MCP; NAZARE sends commands to connected PLCs. Track request status separately from actual outcomes.
ExploreFind what you need now, without leafing through manuals.
Connect relevant material and work context as the starting point for operational agents.
ExploreOne workspace from equipment state to action results.
Review equipment state, AI evidence, work for approval and execution results together. Five equipment demos show how data and operating knowledge connect to responses.
ExploreFactory data. Put to work on your shop floor.
Connect precise equipment signals, production and quality records, and expert knowledge in your own data infrastructure. Start with live monitoring and historical investigation, then build toward AI and autonomous manufacturing with NAZARE.
ExploreDifferent names. Shared meaning.
Manage shared terms and relationships so different tag names do not force you to restart each analysis. AAS and MDR organize the meaning of equipment assets and their data.
ExploreFind the cycle that behaves differently.
Define each production cycle and compare its signals with a reference run. Find changed segments in repetitive production to prioritize quality checks and equipment inspection.
ExploreFind the evidence behind “What changed?”
Compare pressure, temperature, current and other signals around an event. Inspect a short interval in detail or aggregate a longer period to narrow down possible causes.
ExploreDifferent machines. One place for their data.
Connect PLC and CNC signals with records from existing production and quality systems. Choose the fields, verify incoming values, and reuse the same data for monitoring and analysis.
ExploreSee how collection and storage are running.
Manage connected sources, storage capacity and processing status together. Use common screens and policies as the number of machines and operating tasks grows.
ExploreGive every reading its manufacturing context.
A pressure reading is useful when you know which machine and measurement point it belongs to. Link plants, equipment, products and signals so people can understand them and AI can reference them.
ExploreKnow the condition of the data you rely on.
Distinguish missing, late and out-of-range values. Manage original records alongside processing criteria to improve confidence in data used by engineers and AI.
ExploreMake recurring data preparation repeatable.
Clean values by condition and create calculated signals. Prepare equipment readings and existing system records for their intended use, reducing repetitive work before analysis.
ExploreTurn stops and replacement records into improvements.
Record why a machine stopped, not just how long. Review recurring downtime reasons and parts replacement history to identify opportunities to reduce production losses.
ExploreKeep factory data on site. Control who can use it.
Operate on premises when factory data cannot leave the site. Assign equipment and functions by responsibility and separate data access from control permissions.
ExploreMore machines. More history. Find the moment that matters.
High-resolution signals from hundreds of machines quickly become a large history. NAZARE keeps that data on site and queries the equipment, time windows and fields needed for live use and long-term analysis.
ExploreFeed live, precise signals into anomaly detection.
Process-specific ML needs precise data at the moment of inference. NAZARE prepares historical training data and live signals; NAVIS combines results with site knowledge to explain findings and plan responses.
ExploreTurn expert know-how into data AI can use.
What is abnormal, what should be checked first, and what action is appropriate? Ingkle specialists capture this knowledge through interviews and tools, linking equipment, parts, signals, criteria and actions. This is manufacturing ontology.
ExploreSee which machine needs attention now.
Move from plant-wide operating states to individual machine signals. See running, waiting, stopped and alarm conditions together to reduce time spent checking machines in person.
ExploreTurn expert know-how into data AI can use.
What is abnormal, what should be checked first, and what action is appropriate? Ingkle specialists capture this knowledge through interviews and tools, linking equipment, parts, signals, criteria and actions. This is NAZARE ontology.
ExploreFollow data from collection to use.
Design data flows and ownership around sampling, retention, queries and AI workloads.
ExploreCan your data answer the question you care about?
Assess whether connected data is ready for AI across six practical dimensions.
ExploreUnfamiliar terms, explained simply.
Understand platform terminology in the context of equipment data and operational work.
ExploreMeasure improvement using consistent criteria.
Compare consistent conditions and denominators, separating usage from operational outcomes.
ExploreA practical way to organize site knowledge.
Start with the entities, relationships and rules needed for one recurring question.
ExploreDecide what the first trial should prove.
Design a focused evaluation of operational usefulness, not connectivity alone.
ExploreInjection molding · Same recipe. Why are more parts failing?
Setpoints do not tell the whole story. Link mold temperature, cooling and filling behavior to inspection results from the same cycle.
ExploreInspection system · More defective parts—or a change in inspection conditions?
More rejects do not automatically mean the process should change. Link image quality, inspection settings and upstream production by part identity.
ExploreMachining center · Changing tools too early—or too late?
Tool age alone misses actual condition. Connect load and vibration from the same cutting segment with tool history and measured part quality.
ExplorePress · Higher force. A die problem or a material change?
One force alarm is not enough to decide what to check. Compare good cycles, die usage and material lots for the same product.
ExploreManufacturing robot · Same weld path. Why does quality change?
Connect data from robots that make products through welding and assembly. In this welding example, link current, voltage and part positioning to inspection results to locate affected welds and prioritize checks.
ExploreFind your industry’s starting point for manufacturing AI.
Semiconductors, displays, automotive, electronics, defense and national research institutes. Combine a shared data foundation with each site’s process knowledge.
ExploreSpend less time gathering experiment data. Focus on research.
Connect specialized instruments, test equipment and experiment records. Build a secure environment for large volumes of raw data, comparative analysis and AI research.
ExploreWith Ingkle, focus on the services and solutions that set you apart.
Equipment connectivity, storage, queries and scaling recur at every customer site. Use NAZARE as shared data infrastructure and focus on the business features, analytics and AI services your customers need.
ExploreAnswers before you get started.
Review connectivity, product scope, existing systems, AI adoption and operations.
ExploreWhat server capacity does your site need?
Size for tags, rates, retention, concurrency and availability, not equipment count alone.
ExploreChoose a product configuration for your needs.
Choose a scope from data collection through operational analysis and AI agents.
ExplorePlan for operation after deployment.
Choose support for current-version stability or continued product improvements.
ExploreWhen your process needs a tailored prediction model.
Scope specialized prediction or classification models after reviewing data and validation requirements.
ExploreMake the initial scope clear.
Separate product supply, integration, data preparation and operating ownership.
ExploreStart with one problem before changing the whole factory.
Move through discovery, focused validation, operational handover and expansion.
ExplorePrepare for manufacturing AI with national and local government programs.
Explore Ingkle’s national R&D, manufacturing AI smart-factory and Cheonan–Chungnam programs, and identify an adoption path for your factory.
ExploreStart together
Tell us what you want to improve first.
Tell us about your equipment, data and priorities to review a path from connection to operations.