Complex system design and evaluation

Since Sherpa Engineering’s creation in 1997 we have participated to innovative projects, jointly with our industrial customers, to design complex systems that include controls.

From our long experience we have defined and promote among our customers our own design methodology for controlled systems that is being implemented in the PhiSystem product. PhiSystem is the result of a common laboratory with CEA-LIST starting in 2012.

Our activities

Our approach is particularly adapted to mechatronic systems and more generally to controlled systems that assemble manufactured components (mechanics, hydraulics, electrical, thermal, etc), electronic components (instrumentation) and information components (communication, control, decision).

In our model oriented approach, the controlled system design requires modeling and simulation to evaluate design concepts and to make the best choices as soon as possible during the development phase.

Our modeling activity is systemic, meaning that the system is considered globally and not the sum of its parts. Our methodology is used to:

  • Requirements definition and traceability at global level including architecture, requirements, behavior and tests.
  • Simulation model definition and specifications: global model of the system and its environment
  • Control definition and specifications: global modeling of physical, information and decision components in order to define the architecture and the control function
  • Knowledge management: the systemic model is used to collect project related model libraries and technical documents

Methods and tools

Our methodology for controlled system design is based on the General System Theory (GST) and is part of the general System Engineering process. More precisely our systemic modeling methodology verifies the following properties:

  • General system: system is structured, active, evolving and is conceived within its environment
  • Controlled system: system includes the control and the users
  • Multi-level: nature of the system varies depending on the system level
  • Multi-facets: the systemic model includes all facets (requirements, architecture, behavior and tests)
  • Multi-criteria: the systemic model is multi-criteria to be able to integrate all stakeholders’ points of view
  • Simulation: the model allows to simulate specifications, in other words, to validate the system at all levels
  • Collaborative work: methodology uses natural concepts and a standard language that encourage exchanges among all project contributors

Our tool, Phisystem, implements our methodology over an open source tool (Papyrus) and the standard modeling language SysML.


We propose the following training modules (for more information go to training page):

Code Description Duration


System approach and model-oriented engineering

3 days


Design of mechanical systems with loop control

3 days

Key commercial references:

Year Customer Description Manpower



Requirements engineering and control system modeling for a turbo machine

15 persons



Control system of a life support system for a long time spatial mission

2 persons



Modeling methodology to design the architecture of an electrical vehicle

1 person.month


Alstom Transport

Methodology for a control design of an air-conditioning system

1 person.month


Schneider Electric

Model-based design workflow

3 person.month



Modeling platform and evaluation of an electrical vehicle architecture

6 person.month