Væringjar II / Independent Research Programme · Kyiv, Ukraine

Control after failure.

A new control methodology for cyber-physical systems aimed at preserving the highest achievable level of useful operation after failures and irreversible damage.

Scientific foundation →Implementation roadmap

CURRENT STATE: SCIENTIFIC FOUNDATION + REFERENCE TECHNOLOGY UNDER DEVELOPMENT

Methodologysurvivability · residual resources · adaptive control
SoftwareC++ → ROS 2 reference implementation
SiliconRISC-V subcore → SoC IP exploration
Validationrobotics · aerospace · supervised research
Research objective

Maximise survivability after the system has already been damaged.

The controlled object may be irreversibly changed by a failure. The research does not assume restoration of the plant. It develops methods to estimate residual resources, determine the remaining controllable functionality and establish a new control strategy for the damaged cyber-physical system.

\[ \mathcal P_0 \xrightarrow{\;\mathcal D\;} \mathcal P_d, \qquad \boldsymbol{\rho}(t)\in[0,1]^n, \qquad c^\star(t)\in \operatorname*{arg\,max}_{c\in\mathcal C_{\mathrm{feas}}(\boldsymbol{\rho}(t))} S\!\left(c,\boldsymbol{\rho}(t)\right). \] Damage changes the plant. Control is reformulated around the resources and functions that remain.
Research state

The methodology and its implementation are being built together.

01 · ACTIVEScientific methodologySurvivability metric, residual-resource model, configuration selection, control reformulation and stability conditions.
02 · ACTIVEReference softwareIndependent C++ classes → ROS 2 nodes, packages and components.
Research architecture

Mission → Function → Service → Resource.

The programme links resource health to feasible functions, mission priorities, configuration choice and a stability-checked control transition.

Architecture →
POST-FAILURE CONTROL LOOPunder development
01
Damage / degradationThe physical object and available control authority change.
02
Residual resource state \(\boldsymbol{\rho}(t)\)Estimate what resources remain and at what level.
03
Configuration selectionEvaluate feasible post-failure configurations.
04
Stability gateReject transitions that do not satisfy the required stability condition.
05
Continued controlOperate the damaged plant inside a reduced verified envelope.
Exploration domains

We need hard systems, not easy demos.

01 / ROBOTICS

Robotic systems

Manipulator and mobile-platform cases where actuator, sensor, compute or communication loss changes the achievable control problem.

Research track →
02 / AEROSPACE

Aerospace & unmanned systems

Flight-control cases where degraded actuation, sensing, power or communication still leaves residual control authority.

Research track →
03 / SUPERCRITICAL CPS

Other critical systems

Systems where immediate shutdown is not always the safest or operationally acceptable response to partial failure.

Open problem →
Why this site exists

Make the research traceable while it grows.

01
Publish the scientific basis

Definitions, models, equations, assumptions and peer-reviewed outputs.

02
Expose implementation progress

Software architecture, test gates, experimental artefacts and later hardware prototypes.

03
Connect supervised research

Selected problems are developed through PhD and MSc research work under Dmytro Humennyi's supervision.

04
Build credibility before commercialisation

The current priorities are scientific quality, implementation evidence, reproducibility and serious research partnerships.

Open dialogue

Research question, implementation idea, laboratory or validation case?

Væringjar II is open to scientific dialogue, critique, joint experiments and long-term research partnerships. There is no commercial offering at this stage.

Collaboration →