Research
Find the variables that truly shape system behavior, separating structure from correlation and noise.
INDEPENDENT R&D / SYSTEM EXPLORATION
Explore the impossible.
Build answers that run.
I research complex systems, real-time infrastructure, autonomous control, and intelligent decision systems. I start with problems that are not yet clearly defined, then turn uncertainty into working systems through experiments, engineering, and continuous validation.
CURRENT STATUSExploring systems beyond established boundaries
STATEEXPLORING
BOUNDARYUNKNOWN
PROTOTYPEACTIVE
ITERATION0042
01 / POSITION
I work where answers are incomplete.
Problems with standard answers rarely hold my attention. I am drawn to incomplete data, unclear boundaries, and systems where existing methods no longer fit—especially when solving them could redefine what the system is capable of.
Engineering
Turn research into systems that are operational, observable, maintainable, and built to last.
Exploration
Create new possibilities through prototypes and continuous validation when no reliable path exists.
02 / RESEARCH FIELDS
How the unknown becomes a system.
Tools change. What endures is a rigorous understanding of state, feedback, constraints, and failure modes.
01
Complex Systems
State, feedback, causality, and emergent behavior.02
Real-time Infrastructure
Asynchronous runtimes, state synchronization, low-latency processing, and reliable execution.03
Autonomous Systems
Systems that sense, decide, act, and continuously correct their own behavior.04
Human × Machine
New forms of collaboration between human judgment, machine computation, and automated action.03 / SELECTED EXPERIMENTS
Every system begins as a question.
These records include successful results, failed assumptions, boundary conditions, and questions that remain unresolved.
EXP / 01
Autonomous Thermal Control
How can a multi-sensor system remain stable across noise, hysteresis, hardware failure, and continuous operation?
ACTIVE
EXP / 02
High-density Event Reconstruction
How reliably can system state transitions be reconstructed when information is missing, timestamp precision is limited, and event order is incomplete?
EVOLVING
EXP / 03
Long-running Reliable Runtime
How do we build low-dependency automation that is recoverable, observable, maintainable, and able to operate continuously in real environments?
ACTIVE
EXP / 04
Context-aware Decision Systems
How can a system recognize its operating context and stop applying one fixed decision process to every environment?
ARCHIVED / LEARNING
04 / METHOD
An R&D process designed for uncertainty.
Research is not the search for evidence that supports an idea. It is the fastest honest path to discovering why the idea may be wrong.
Observe
Study the real system instead of starting from a conclusion.
Decompose
Separate data, state, constraints, and feedback.
Prototype
Test the critical assumption with the smallest viable system.
Measure
Define repeatable experiments and explicit failure criteria.
Refine
Remove ineffective complexity and preserve essential structure.
Operate
Let the system keep learning in the real environment.
05 / PRINCIPLES
Innovation should not come at the cost of reliability.
The closer a system moves toward an unknown boundary, the more it needs rigorous evidence, explicit failure criteria, and maintainable engineering foundations.
Evidence over narrative
Evidence outranks a persuasive story.
Systems over features
System capability matters more than feature accumulation.
Simplicity after understanding
Simplicity is what remains after unnecessary complexity is understood and removed.
Failures are data
A failed experiment still reduces the space of the unknown.
Reliability is innovation
Long-term stable operation is itself a form of innovation.
Boundaries are temporary
A boundary is often a problem that has not yet been redefined.
06 / CURRENT EXPLORATION
Boundaries currently in motion.
- 01Autonomous runtime recovery
- 02Event ordering under incomplete information
- 03Context-aware adaptive control
- 04Long-horizon system reliability
- 05Machine-assisted research workflows
I believe meaningful R&D does not reproduce answers that are already known.
The impossible is often just
an untested system.
What appears impossible is often a system that has not yet been tested correctly.
07 / NEXT UNKNOWN
Build beyond the known.
If you are investigating a problem without a standard answer, or building a system that existing tools cannot yet deliver, our paths may be worth crossing.
admin@mifen.me