TwinSpace

The aim of the lighthouse project is to spread the hinge function of digital technologies between economic and ecological benefits. Specifically, this is to be achieved by optimizing (safety-relevant) software for embedded systems in terms of resource requirements (e.g. CPU, GPU, runtime). To this end, automated reverse engineering of compiled code is being pursued. This code is then simulated on various emulated hardware (HW) platforms and the resource requirements are measured. As soon as the appropriate HW platform is selected, the code is recompiled for the corresponding HW platform. Reverse engineering poses no cyber risk, as only non-functional aspects are transferred. Traditionally, resource optimization for software (SW) is ignored, as the speed of HW doubles every two years anyway. By means of the developed, tested and applied green coding development tools with optimized resource management - TwinSpace - the SW development in the (safety-relevant) mass business of embedded systems is to be revolutionized. This will optimize both economic aspects, for example through more efficient development processes, and ecological aspects through increased energy efficiency or the reduction of CO2 through smaller chip areas. The developments are tested, applied and then rolled out on a broad scale.

Our aim is also to use simulators for ESL (Electronic System Level) for load profile generation and evaluation. These simulators offer significant advantages in the development of a new application consisting of hardware and software, as they can replace the hardware in early development phases. It can be distributed as a software environment and is available as early as possible for the development of the application software. This allows the hardware system and the hardware/software combination to be tested at an early stage, thus avoiding the production of early and potentially faulty prototype systems.
An ESL simulation environment can be used directly for the development of the application software. This speeds up the development cycle considerably. The integration of load profile generation means that software twins are available for hardware design at an early stage for the purpose of realistic testing of the overall environment. This again reduces the need for physical prototypes and speeds up the development process. The interaction between hardware and software is thus simulated and broken down into pure software. The use of realistic load profiles also ensures that development partners do not have to disclose any internal, potentially protected (program) data during the development process, as the load profile has an input/output or bus access and communication behaviour that is equivalent to the target application.

Funded by BMWK (GreenTech Innovationswettbewerb); now listed as BMFTR, GreenTech Innovationswettbewerb (Digitale Technologien), grant 01MN23013H (Universität Rostock sub-project).