Thorough application of Scenario Management requires the concerted, skillful, and interconnected application of many sophisticated technologies and methods.
Artificial Intelligence
is used to
- to assure excellent functionality in control,
- for Scenario Mining by Machine Learing / Data Mining tools,
- to assure data quality with according expert systems
by Artificial Neural Networks, Machine Learning Methods, Expert Systems, formal methods.
Big Data Analytics
is used
- to manage the huge amount of scenarios and data in the catalog
by appropriate tools.
Change/Anomalies Detection
is used
- for quality assurance of the data,
- to identify critical situations in operations,
- to identify new situations and ODD exceedance
by Expert Systems, unsupervized Machine Learning Methods, classical signal processing analysis.
Complexity Management
is used
- to control the complexity of the requirments,
- to assure system robustness,
- to avoid unnecessary development efforts
by clever system design and architectures, by modern complexity rating methods.
Conditional Probabilities
are used
- to handle the stochastic nature of the problems,
- provide best system performances,
- assure effective experimental design
by Bayesian approaches.
Conflict Analysis
is used
- to identify and resolve functional requirment conflicts quickly,
- to reduce the main source of AI malfunctions,
- to reduce the complexity of the functional requirements,
- to improve system robustness
by data-based requirements represenation and special, sophisticated metrics.
Effectiveness & Impact Assessment
is used
- as premium system validation method,
- to identify the components with best/least effect on system performance,
- to steer development efforts
by skillful systems engineering.
ODD Management
is used
- for safe system operations,
- for system explainability,
- for effective test design and validation
by skillful system engineering.
Quantitative Risk Assessment
is used
- for risk estimation,
- for estimating impacts in case
- of system or component failure,
- of system misuses
by stochastic simulation methods.
Reference Functions & Architectures
are used
- for system and component ratings and assessment,
- for functional decomposition to break the curse of dimentionality
by clever design.
Robustness Management
is used
- to prevent system malfunction,
- to validate system functionality
by Stochastic Simulation.
Sampling Techniques
are used
- for efficient robustness management
- for effective design exploration
- effective test design and validation
by Design of Experiments and Monte Carlo samplings.
Stochastic Simulations
is used
- for realistic simulations
- of human behaviours,
- of scattering operational conditions
by Monte Carlo simulation.
System-of-Systems Engineering
is used
- for functional decomposition to reduce development efforts,
- for modularization of solutions to improve scalability
by clever minds and skillful engineers.
Various Labeling Methods
are used
- to build virtual sensors
- for collision risks,
- for congestion risks,
- for failure risks,
- to generate training data outputs automatically
by Monte Carlo Markov Chains, and equivalent stuff.
Escaping the complexity lock with technology!