Platform Performance Suite

PPS- Platform Performance Suite

The Platform Performance Suite is a performance analysis methodology, with supporting tooling, that aims at providing detailed insight on the software execution of Cyber Physical Systems. The PPS adheres to the Timed Message Sequence Charts (TMSC) modeling formalism. All the PPS visualizations and analysis methods are based on TMSC. Various analysis methods are offered out of the box, like critical path, slack and root-cause analysis. The PPS method has been applied and proved its value on various systems and use cases (Roos, Clearing the critical software path, 2021) (Roos, Taking performance analysis to the system level, 2023).


PPS report - pdf

PPS is published in the open source - read more


PPS can be used to get insight on the detailed software and system execution. The explicit dependencies between software actions facilitate the understanding of the system and software execution.

The framework starts with collecting traces during the system execution. These traces are used to infer models of the system execution behavior. This derived model is then used for getting insight on system execution and analyze the system performance.

PPS method

The overall PPS method as used in brown-field engineering is visualized in Figure 1. The model of execution is reconstructed from execution traces and aligned with the model of the life cycle software architecture, which is inferred from existing design and implementation artifacts. The analysis of the model of execution (reactive) provides the root-cause (artifact) of a performance issue. The analysis of the life cycle software architecture model (proactive) provides with potential performance issues. This model can also be used in a what-if-analysis optimization loop where the artifacts are tuned towards higher productivity.

Figure 2. The PPS method in brown-field engineering.

An important aspect of the PPS framework is the classification of artifacts and modeling elements in different stages during the software lifecycle. PPS considers five different lifecycle stages, namely, the Specification, Implementation, Deployment, Runtime-Instantiated and Runtime-Stimulated. The Runtime Stimulated is the cornerstone of the PPS and adheres to the TMSC modeling formalism (K.Triantafyllidis, Niknam, & Blankenstein, 2023). The TMSC, as well as the other lifecycle stages can be modeled in the PPS tool by the use of dedicated domain specific language(s) and can be also be used as parser for traces that adhere to the TMSC modeling formalism. The TMSC models provide explicit dependencies between software execution entities that facilitate semi-automated performance analysis techniques.

Visualization & Analysis

In the PPS tool, the TMSC models are visualized in Gantt-chart like diagrams interconnected with dependency arrows. PPS provides various analysis types, like critical-path, slack and root-cause analysis. The analysis is applied on the visualized models and shown as overlay on original model, facilitating analysis observations and diagnosis at specific and different abstraction levels.


PPS tool comes with examples of modeling aspects tailored to the well-known CORBA architecture. For the application of the tool the user may need to model various stages of the lifecycle software system architecture. These are typically domain specific concepts and refer to the actual implementation and architecture patterns used for the implementation of the system of interest.

Download and Install

Platform Performance Suite (PPS) (tno.github.io)

Source code
TNO/PPS: Platform Performance Suite (github.com)