Open positions

PhD study candidates

Diagnostics and Modelling of Industrial Ethernet Networks

Industrial Ethernet is getting more and more involved in manufacturing systems to connect the controllers and remote peripherals. The necessity to keep real-time qualities of the communication as well as the requirements to transfer non-real-time data over the same cable bring new challenges. Some of them are addressed in this topic for PhD study (more details are provided here). There is a strong relation to industrial partners such as Skoda Auto, as well as research partners such as TU Dresden.

Summer Internship

For the summer of 2014 students are encouraged to participate in this programme to support research projects being done in the Industrial Automation Group. The following topics are available for individuals or groups of students. It is possible to do the summer internship just for the summer holiday or it is possible to be extended to other projects/courses, in which you receive credits during the studies. For more information contact pavel [dot] burget [at] fel [dot] cvut [dot] cz (Pavel Burget).

Modelling and Simulation of Profinet Communication

This work serves as a preparation for creation of a diagnostic tool usable with Profinet networks in industrial plants. Special focus is on car industry.

  1. Get acquainted with the OmNet simulation environment  
  2. Based on existing model libraries for Ethernet networks create a basic model of Profinet IO Controller and Profinet IO Device for cyclic data communication  
  3. Model various Profinet network topologies with various traffic loads based on real installations

Activities and SW: OmNet, C#(C++)

Diagnostics of Ethernet Communications

BANY Premium is a unique diagnostic tool suitable for traffic sniffing in Profinet networks with precise time stamping and scripting possibilities.

  1. Get acquainted with BANY Premium.  
  2. Study the possibilities of the scripting interface.  
  3. Write scripts how to operate BANY externally such as setting the triggering conditions, generating frames, etc.

Activities and SW: Ruby or C#(C++). If you don't know Ruby, it is easy to learn quickly.

Profinet Topology Visualisation and Modelling

The network topology can be obtained by reading the respective information from the nodes using SNMP. However, there may be nodes, which do not obtain the topology-related information to be read.

  1. Study existing graph-based algorithms for modelling network topology  
  2. Based on the retrieved information build the topology and identify white spaces  
  3. Estimate the presence of devices in the white spaces based on relevant information from the known neighbours  
  4. Visualise the topology in various forms using existing libraries (such as D3), take into account tree and ring topologies and their combinations.

Activities and SW: JavaScript

Profinet Diagnostics Tool Enhancements

Information about the devices in the Profinet networks can be obtained using RPC protocol. The task is to enhance the existing RPC library and to write functions, which read specific data sets from the devices such as those regarding topology, device identification, diagnostics, etc.

  1. Study the current implementation of the RPC library.  
  2. Complete the missing functions to the library.  
  3. Write functions to read the required data sets from the devices.

Activities and SW: Ruby, Ruby on Rails. If you don't know Ruby, it is easy to learn quickly.

Testing of Software

Write tests for the currently available functions for RPC communication, topology discovery, etc., which are parts of the existing Profinet diagnostics tool.

Activities and SW: Ruby, RSpec

Profinet IRT Diagnostics

The purpose is to check the existing Profinet IRT segment and its devices, whether all devices contain relevant scheduling information. Use the RPC library to retrieve the data sets regarding the message schedule, network topology, etc. from the devices and compare it to the devices' parameters obtained from their GSDML files. This work is to be used to test the correctness of the Profinet IRT schedule generated by independent schedulers.

Activities and SW: Ruby, Ruby on Rails. If you don't know Ruby, it is easy to learn quickly.

Improvements of the models in LabLink

For the waterplant, which is used in LabLink, create the following enhancements:

  1. Improve and make reliable the remote manipulations with the model. Analysis of the TCP/IP communication may be necessary. Implement the possibility of remote power-off feature using an IP relay.  
  2. Test different PLCs, which work with the plant, such as CPU 315-2 PN/DP and IPC Microbox. Check the possibilities of the latter stemming from using WinAC and PC-based programming.  
  3. Integrate the acyclic communication to the model. Integrate diagnostics functions and possibility to restart the frequency inverter (acknowledge errors).  
  4. Use the model of the waterplant created in Matlab and prepare a setup in which the PLC control can be done on this simulation model instead of the work with the real plant.  
  5. Improve the documentation of the model.

Activities and SW: Step7, C#, Ruby, Ruby on Rails. If you don't know Ruby, it is easy to learn quickly.