Industrial Agent
Roadmap- Edge intelligence
- Multi-machine coordination
- Dynamic process optimisation
- Diagnostics
- Natural-language interaction
Industrial AI
Industrial AI That Starts With the Process
Not a generic AI platform. Process intelligence, built on real equipment, real process signals and manufacturing know-how.
Our position
Do not build a generic AI platform.Build process intelligence.
Leidao industrial AI grows from real equipment, real process signals and manufacturing know-how.
- 1Physical process
- 2Sensing & data
- 3Process models
- 4Industrial decision
- 5Production feedback
Four stages
- 01In development
Process Modelling
Capture parameters, signals and outcomes from real machining and production, and structure them into models of how a process behaves.
- 02In development
Industrial Skills
Package proven process knowledge into reusable skills that equipment and production lines can call on.
- 03Roadmap
Industrial Agents
Agents that read process state and propose or coordinate actions under engineer oversight.
- 04Roadmap
Industrial Cloud
A shared platform for process data, models and skills across sites.
This is the architecture and development path, not a claim that every function is available. Process modelling and industrial skills are in development. Industrial Agent and Industrial Cloud are roadmap items.
Process modelling
Process Modelling
SPETEC and Leidao design every ultrasonic machine tool as a data-collection terminal, so that real machining becomes structured process knowledge.
Signals captured
- 01Vibration
- 02Load
- 03Thermal deformation
- 04Spindle pull and spindle state
- 05Quality outcome
- 06Tool condition
Every machine becomes a source of process knowledge.
A process-parameter library
The architecture builds libraries of process parameters for difficult-to-machine materials, packaging the best combinations of material, tool and working condition into standard skill packages.
- Titanium alloys
- High-temperature alloys
- Ceramics
- CMC
Industrial skills
From Expert Know-how to Industrial Skills
A skill package turns the hand-feel of an experienced machinist into a repeatable, transferable digital asset.
What a skill package contains
- 01Material
- 02Tool
- 03Machine state
- 04Process parameters
- 05Quality criteria
- 06Constraints
- 07Expert knowledge
Target materials
- Titanium alloys
- High-temperature alloys
- Ceramics
- CMC
Skill packages are in development, starting with the materials above.
Roadmap
Industrial Agent and Industrial Cloud
The next steps build on process models and skills.
Industrial Cloud
Roadmap- Skill library
- API ecosystem
- Tool and material partner integration
Both are roadmap items and are not yet available.
Research
Fractional-Order AI Research
Leidao's intelligent-manufacturing roadmap includes joint research on fractional-order AI for equipment, production-line and factory optimisation.
- 01
Equipment: modelling, self-optimising parameters, diagnostics
- 02
Production line: dynamic scheduling and multi-product changeover
- 03
Factory: coordinated optimisation across lines
Control methods
Fractional-order ultrasonic RTCP
Keeps the ultrasonic vibration direction aligned with the surface normal during five-axis motion, to reduce chatter on complex surfaces.
Fractional phase-locked frequency tracking
Locks onto the resonance point quickly when the load changes.
Fractional PID amplitude control
Closed-loop amplitude control, with a target fluctuation within ±0.5 µm.
Principles
How We Approach Industrial AI
Industrial AI succeeds when it delivers measurable value on a real production line, not when the algorithm is the most advanced.
- 01
Start with the real scene
Value comes from real, measurable business scenes, not from technology for its own sake.
- 02
Build the data foundation
Start from accumulated process data. Without high-quality, standardised industrial data, industrial AI has nothing to stand on.
- 03
Capture expert knowledge
Turn expert experience into knowledge bases and intelligent agents that become lasting enterprise capability.
- 04
Scale step by step
Climb the stairs, do not take the lift. Each level rests on the one below it, with no skipping.
Three engineering requirements
Reliability
Industrial scenes demand precise, stable and verifiable results.
Explainability
Process decisions must be traceable, so engineers can understand and check them.
Data quality
Vibration, temperature and image data must be standardised across systems.
The intelligence loop
From Physical Process to Industrial Intelligence
Industrial AI grows from real equipment, real process signals and manufacturing know-how. The loop returns to production, so every decision is tested on the line.
- 1
Physical process
Cutting, grinding and forming on real equipment.
- 2
Sensing & data
Signals and parameters captured from the process.
- 3
Process models
Process knowledge turned into models.
- 4
Industrial decision
Parameters and actions recommended to the line.
- 5
Production feedback
Results return to improve the process.
- 01In development
Process Modelling
Capture parameters, signals and outcomes from real machining and production, and structure them into models of how a process behaves.
- 02In development
Industrial Skills
Package proven process knowledge into reusable skills that equipment and production lines can call on.
- 03Roadmap
Industrial Agents
Agents that read process state and propose or coordinate actions under engineer oversight.
- 04Roadmap
Industrial Cloud
A shared platform for process data, models and skills across sites.
This is the architecture and development path, not a claim that every function is available. Process modelling and industrial skills are in development. Industrial Agent and Industrial Cloud are roadmap items.
Learn moreEngineering & technology
From Research to Industrial Intelligence
The route SPETEC follows from technology to industrial capability. Each stage shows its own status.
- 01
Research
Materials and process fundamentals.
Technical capability - 02
Engineering
Research turned into engineered solutions.
Technical capability - 03
Equipment
Machine tools and tooling built for the process.
Technical capability - 04
Process
Parameters and know-how fixed into repeatable processes.
Technical capability - 05
Production
Equipment connected into flexible production lines.
In development - 06
Industrial intelligence
Models and skills first. Agents and cloud are roadmap.
In development
A development path. It does not mean every stage is already commercial.
Let's Engineer What's Next.
Tell us about your manufacturing process, equipment or project, or explore a partnership with the group.