Case studies

ELI needed a unified control solution for a new crystallography end station (TREX). The objective was to deploy the open-source MXCuBE beamline software and integrate it with several specialized devices: a STOE StadiVari goniometer, an XYZ SmarAct micro-positioning stage, and an Arinax OAV B-Zoom optical microscope. Together, these components were to enable precise sample alignment and coordinated motion and imaging during laser-driven X-ray crystallography experiments.

The project involved several challenges:

• Remote development constraints
  The project was carried out entirely through remote access. The diffractometer is housed in a clean room environment where numerous risk factors are present, including high-power lasers, soft and hard X-ray radiation, toxic chemical gases, and other hazards. Due to these conditions and the resulting limitations on internet access, remote development proved to be the safest and most effective approach.
• Multi-device integration
  None of the TREX devices were natively supported by MXCuBE, requiring the development of custom hardware interfaces. Each device used different communication protocols and offered limited documentation, increasing integration risk.
• Algorithm complexity
  The client required advanced features such as 3-click centering – a semi-automated method for precisely centering a crystal using image-guided movements. While the algorithm existed in MXCuBE, it was complex, poorly documented, and not directly compatible with the TREX geometry.
• Access and timeline constraints
  Hardware access was limited due to on-site restrictions, delaying early testing and forcing part of the development to proceed using simulations and incomplete information. At the same time, the project required full documentation, training, and formal acceptance tests within a fixed schedule.

S2Innovation delivered a customized MXCuBE implementation tailored to the TREX end station.

• Customized hardware control
  Dedicated drivers and adapters were developed for each device, enabling full control of the XYZ positioning stage, precise goniometer rotation, and live microscope imaging. All devices were integrated into a single MXCuBE interface, allowing scientists to control positioning and visualization from one application.
• Advanced workflows and feature extensions
  The system was extended to support continuous scans, region-of-interest selection, exposure control, and automated scan queues.Mesh scanning was not implemented due to the narrow scan range of the SmarAct positioning stage.
  The critical 3-click centering functionality was adapted by introducing an additional calibration parameter to compensate for orientation mismatches between camera view and motion axes. After tuning, the centering routine reliably positioned samples at the beam center.
• Deployment and testing
  The solution was containerized using Docker to ensure consistent deployment on ELI’s infrastructure. Integration was validated through two on-site test campaigns. Issues identified during the first test were resolved through configuration and code updates, and the second test confirmed full functionality under real operating conditions.
• Training and documentation
  Comprehensive technical documentation was delivered, including in-code documentation and deployment guides. A dedicated hands-on training session ensured that ELI staff could confidently operate, maintain, and extend the system.

By October 2025, the integrated MXCuBE solution successfully passed all verification steps and achieved final acceptance. The TREX end station gained a unified, reliable control system capable of precise sample centering and fully automated scanning workflows.

The adapted 3-click centering algorithm performed consistently and resolved issues observed during early testing. All project objectives were met, and the client reported no outstanding issues or change requests after delivery.

The project demonstrated S2Innovation’s ability to adapt an open-source scientific platform to complex, non-standard hardware environments. Close collaboration with the client and iterative on-site testing proved essential to achieving a robust and user-friendly solution, strengthening the foundation for future cooperation with ELI.

Got a similar concern? Set up a free consultation.

Book a meeting

We developed a dedicated Lima/Tango plugin for ESRF that enables control and data acquisition for the atypical XH linear detector, despite the lack of an SDK and limited access to the hardware. The solution was integrated with the ESRF ecosystem (LimaCCD/Bliss) and put into production, improving measurements on the ID24 beamline.
The project delivered for the European Synchrotron Radiation Facility (ESRF) falls within the domain of so-called Big Science, where research requires advanced infrastructure and unique technological solutions. In 2023, during a six-month collaboration, the project team faced the task of developing a control and visualization system for an exceptional 1D detector. Unlike standard LimaCCD implementations, which typically address 2D detectors, the solution had to be tailored to the specifics of a one-dimensional device.
The main goal of the project was to implement software that would allow convenient and intuitive operation of the new detector, ensuring its use is unified with other devices functioning in the Tango Controls ecosystem. Achieving this required creating an intermediary module between Tango Controls, the LimaCCD library, and the device software, so that operators could use a consistent interface regardless of the detector’s specifics.

The client was ESRF — one of Europe’s largest research institutions specializing in the use of synchrotron radiation. The organization supports international scientific teams by providing tools for research in fields such as physics, chemistry, biology, and materials science. Deploying the new system had a direct impact on operator efficiency and represented an important step in further improving ESRF’s advanced research infrastructure.

During project execution, several difficulties were identified that significantly affected the implementation process:

• Adapting LimaCCD to a 1D detector — standard use cases for this solution mainly involve 2D detectors, so an approach supporting the atypical one-dimensional case had to be designed.
• Unique nature of the device — the detector did not have a standardized SDK (Software Development Kit), and there were no ready-made communication or control tools available. This required the team to develop a custom integration layer.
• Limited hardware availability for testing — due to the conditions under which the detector could be used, software testing required special organization and was not always possible in full scope.
• Lack of test procedures — the absence of standardized scenarios to verify correctness required additional effort to define evaluation criteria and validation methods.

1. Technologies, tools, methods
   The project used a set of tools enabling efficient implementation and integration of the control system. The key role was played by C++ and Python, along with the Linux environment. For detector operation, the LimaCCD and Tango Controls frameworks were used, which form the foundation of the existing control architecture at ESRF
2. .Main challenges and how they were solved
   A key challenge was designing the control code so that it would work correctly with a 1D detector, even though LimaCCD is by default used in 2D configurations. This meant modifying the application-side communication and data-processing logic so the system could correctly interpret one-dimensional data and expose it to operators in a unified way.
   The lack of a standard SDK and ready-made control tools made it necessary to create a proprietary communication module. The team developed data-exchange mechanisms that enabled operation of the detector’s key functions while maintaining compatibility with Tango Controls and the existing control environment.
   Another significant obstacle was limited detector availability for testing. The detector could operate only under specific conditions, which reduced the number of possible verification sessions. To make the most of each session, the team prepared detailed test scenarios and executed the tests as comprehensively as possible during every available session.
   In parallel, test procedures had to be designed from scratch. The team developed a set of scenarios tailored to the specifics of the 1D detector, allowing verification of correct data acquisition, communication stability, and integration with the Bliss data acquisition system used at ESRF.

1. The project’s greatest success
   The most important achievement was delivering a solution that turned a seemingly difficult and atypical detector into an operator-friendly tool. The software reduced task execution time and significantly lowered the risk of user errors. The project demonstrated that even for devices lacking standard control tooling, it is possible to build a coherent and reliable system that meets the requirements of top-class research environments.
2. Impact on the client relationship
   Deploying the system strengthened ESRF’s trust in the delivery team, confirming its competence in working with advanced technologies. The client gained confidence that entrusted tasks can be executed effectively despite atypical conditions and hardware limitations. The collaboration — built on flexibility, good communication, and readiness to solve non-obvious problems — created a solid foundation for further joint projects.

Advice for other clients:

• Define test cases early — precise verification scenarios help plan implementation better and avoid delays.
• Consider the hardware operating environment — the conditions under which the device works may require adjusting test schedules and adding supporting tools.
• Look for flexible communication solutions — regardless of whether the manufacturer provides an SDK, it is worth planning to implement proprietary control modules and integrate them with the existing architecture from the start.
• Ensure system compatibility — early verification of integration with existing tools such as Tango Controls or data acquisition systems helps avoid costly rework.

The ESRF project shows that even highly atypical devices can be incorporated into a unified control ecosystem. Clients choosing similar implementations gain higher operational efficiency, lower susceptibility to user errors, and greater consistency in research team workflows. The benefits go far beyond automation itself — they open the door to faster, more accurate, and more reliable experiments.

Got a similar concern? Set up a free consultation.

Book a meeting

The project carried out in 2022 focused on defining a comprehensive set of graphical guidelines for the UNICOS framework, used at CERN to build supervision applications for industrial processes. According to CERN’s requirements, the guidelines needed to unify the visual identity of UNICOS interfaces, introduce a coherent icon library, define a color palette aligned with ANSI/ISA-101, specify typography rules, and prepare layout templates for key types of UNICOS windows. 

About the client 
CERN is the world’s largest particle physics research center, operating a vast portfolio of technical and industrial systems. The project was part of a broader initiative to modernize UNICOS, a framework that has evolved over more than 15 years and therefore requires visual consolidation and standardization. 

Modernizing UNICOS’ visual identity posed several challenges related to framework expectations, long-term use, and compliance requirements. UNICOS is a CERN-developed controls environment with predefined window types, established panel structures, references to ISA-88 and ISA-101 standards, an extensive library of around 150 existing icons, and strict expectations regarding compatibility with WinCC OA. Over the years, the framework has been extended by many contributors, which resulted in a heterogeneous visual landscape — non-uniform iconography, varied color schemes, and inconsistent layout structures. In addition, the guidelines had to fully comply with ANSI/ISA-101 standards, ensuring proper ergonomics for technical operators and alignment with CERN’s operational practices. Although WinCC OA is a flexible platform, the new guidelines also needed to reflect the technical structure of UNICOS templates implemented on top of it, account for compatibility across Windows Server 2022 and RHEL 9, and consider implementation constraints related to XML-based panel files and CTL scripts. 

Technologies, tools, and methods 
We used WinCC OA as the reference platform, Qt-based tooling for interface design, vector-based tools for SVG/PNG icon creation, and ANSI/ISA-101 and ISA-TR101 as the guiding standards. 

Main design challenges and solutions 
We unified the existing UNICOS iconography, developed an ISA-101-compliant color palette, standardized UNICOS window layouts, and prepared implementation guidelines ensuring future maintainability and consistency. 

Biggest project success 
The key achievement of the project was creating a modern, unified visual foundation for UNICOS that consolidates previously heterogeneous elements, aligns with the ISA-101 standard, integrates smoothly with WinCC OA’s technical environment, and can guide CERN’s future UI development for years to come. 

Impact on the client relationship 
The project delivered a clear and coherent graphical standard, simplified paths for future HMI modernization, improved consistency across UNICOS applications, and a reusable design system compatible with CERN’s operational needs. 

Consider framework specifics from day one, prioritize clarity and usability, design reusable components, and maintain transparent communication. 

Got a similar concern? Set up a free consultation.

Book a meeting

The aim of the project was to design and implement an advanced control system for the ForMAX beamline at the MAX IV synchrotron in Sweden—one of the most modern research facilities in the world, specializing in producing synchrotron radiation of exceptional brightness. The facility includes two storage rings (1.5 GeV and 3 GeV) and state-of-the-art beamlines tailored for research in chemistry, physics, structural biology, and materials science. 

ForMAX supports, among others: 

• 3D imaging of porous material structures at micrometer scale with 1-second time resolution, 
• studies of bio-based nanostructures with millisecond resolution. 

Our team was part of an interdisciplinary project group, working with partners on a system that met ForMAX’s demanding requirements. We implemented a flexible and reliable solution enabling precise control of experimental parameters (position, energy, beam intensity, sample alignment) and real-time data processing. The system supports step and continuous scanning, full hardware–software synchronization, and the integration of tomography with SWAXS techniques in a single instrument. 

The ForMAX beamline project at MAX IV involved a series of complex technological and operational challenges. The client—a multidisciplinary team of scientists and engineers—expected a system that would meet rigorous requirements of precision, flexibility, and reliability in the context of synchrotron experiments. Ensuring seamless integration of diverse devices and technologies within a single control platform was crucial. 

Technologies, tools, methods 
Building the control system required solutions tailored to the specific experimental conditions where both precision and adaptability to changing parameters are critical. The implementation team had to account for the diverse nature of devices, dynamic operational loads, and the need for remote control and visualization.

Key challenges solved: 

• Precise equipment control 
  The software handled numerous mechanical and optical devices—including linear motors (IcePAP), movable sample stages, monochromators, detectors, and beamline components. High positioning accuracy was achieved using pseudomotors and dedicated macros, enabling synchronized movement of multiple components in real time. 
• Integration of multiple technologies 
  The system integrated distributed solutions using TANGO Controls and Sardana frameworks, enabling client-server communication between subsystems. The software was extended to support non-standard devices such as PandABox (for trigger pulse generation) and trajectory controllers (for parametric scans). 
• Step and continuous scanning support 
  The system implementation included various scanning modes, such as point, line, and continuous mesh scans. A key requirement was enabling trigger-based measurements without stopping motor movement. This was achieved through integration with PandABox and adapting the synchronization algorithm for pulses. 

During the implementation of the control system, several significant technical issues were encountered, which required advanced analysis and modifications to the system’s operating algorithms.

• Drift during continuous scanning 
  The issue was caused by PandABox generating extra synchronization pulses. This stemmed from incorrect formulas in Sardana, which miscalculated total scan time and motor velocity, resulting in an extra pulse per scan line. The solution involved overriding Sardana’s default formulas for velocity and scan time calculation in the meshct class, ensuring position stability and correct synchronization. 
• Delays between scan lines 
  Problems arose during continuous mesh scanning with active PandABox synchronization. Detectors were being armed at the start of every scan line, introducing delays. This was optimized by modifying the meshct class to arm detectors once at the beginning of the scan instead of at every line. Result: delay reduction from 3.04 s to 0.732 s per line. 

Performance and usability improvements 
The most important achievement was a significant improvement in ForMAX’s control system performance. The implemented solutions allowed synchrotron experiments to be performed faster, more precisely, and more reproducibly. By optimizing the scanning process, we reduced measurement delays, minimized synchronization errors, and improved real-time data acquisition and processing. End users gained a more intuitive work environment and a stable, reliable system for conducting complex experiments. 

Collaboration and relationship building 
The project earned client recognition not only for the quality of the technical solution but also for the partnership-driven approach of the implementation team. Collaboration with scientists, engineers, and technicians at MAX IV was based on regular communication, problem-solving flexibility, and readiness to adapt the system to real user needs. The outcome was strengthened client relationships and trust that may lead to future joint R&D projects. 

Based on the ForMAX experience, we recommend several key actions for organizations planning to deploy advanced control systems in research and industrial environments: 

1. Engage technical teams and end users early. This helps identify real operational needs and design a system tailored to the application context.
2. Analyze and test synchronization thoroughly. Device communication and synchronization must be validated in near-real conditions—especially for continuous scans requiring precise triggering between motor movement and data acquisition. Early testing helps prevent potential quality or continuity issues. 
3. Leverage open, modular architectures. Frameworks such as TANGO Controls and Sardana provide a solid foundation for future development. Their flexibility facilitates system expansion, integration of new components, and long-term maintenance. Open-source architectures also allow modification or overriding of base components—as in the case of the Sardana meshct class—to meet specific experimental requirements. 

Got a similar concern? Set up a free consultation.

Book a meeting

The project was born out of the need to create a versatile and precise tool for X-ray emission spectroscopy (XES), enabling work with samples in various states – liquids, powders, and thin films. A von Hamos geometry was used to combine data acquisition and analysis in a single integrated system. The goal was to build a complete working environment encompassing both the measurement process and analytical workflows within one application.

Two main components were developed as part of the project:

• A Tango Device Server for the Andor Newton camera
• A data analysis module with a set of image-processing algorithms, all enclosed in an intuitive graphical interface.

Implementing the integrated system delivered tangible results that significantly improved the team’s daily work:

• 60% faster access to ready experimental results – automation and the reduction of manual steps drastically shortened the wait time for outcomes.
• One unified environment from measurement to analysis – users no longer need to switch between multiple tools, which reduces errors and simplifies usage.
• Intuitive GUI reducing user onboarding time – the user-friendly interface helps new team members start working faster without lengthy training.
• Process automation – fewer errors, more repeatability – the system took over many manual and repetitive tasks.

Improved collaboration between research and technical teams – a standardized environment enhanced communication and supported efficient achievement of shared goals.

Throughout the project, several key challenges that had previously hindered efficient experimentation and data analysis were addressed:

• Manual and time-consuming data processing – previously, data was processed using external scripts and tools, requiring significant user effort. The client relied mainly on manufacturer-provided software and custom scripts in Mathematica and Python.
• Lack of integration with Tango Controls – limited automation – earlier tools did not support direct cooperation with the control system, making it difficult to automate acquisition and analysis.
• Inconsistent working environment and frequent export/import errors – switching between various programs and formats caused errors and inefficiencies. Manual transitions between tools introduced chaos and reduced workflow effectiveness.
• Inefficient algorithms for large datasets – the previous analysis methods were too slow for high-resolution data, limiting the team’s ability to process a higher volume of samples.

The designed solution combines precision, automation, and ease of use, creating a complete environment for spectroscopic data acquisition and analysis. Every stage – from image capture to final processing – was integrated into one coherent system. As a result, users benefit not only from faster operation but also from full control over the experimental process.

Core system components include:

• Device Server for the Andor Newton camera (PyTango) – enables control over camera parameters and acquisition modes.
• Analytical module – a set of image processing algorithms: Gaussian filter, median filter, Savitzky-Golay smoothing, energy calibration, and center of gravity (COG) calculation. These algorithms help reduce noise, smooth data, precisely locate spectral lines, and convert raw data into a format ready for analysis.
• Intuitive GUI – developed in PyQt and Taurus, providing a clear and integrated interface for both acquisition and analysis.
• Optimized batch processing – vectorized operations using NumPy enhanced performance, particularly for high-resolution datasets where previous methods fell short.
• Full integration with Tango Controls – allows for automatic process control and synchronization.

The system’s implementation brought not only technical improvements but also strengthened the partnership with ELI Beamlines. The team appreciated our openness, flexibility, and responsiveness to evolving needs.

During the project, we overcame several obstacles, including the lack of documentation for the Andor Newton camera and legacy analytical methods. Through reverse engineering and close collaboration with the client, we were able to reconstruct, optimize, and tailor the necessary algorithms to fit ELI Beamlines’ infrastructure.

The result is a system that is technically refined and user-oriented – intuitive, flexible, and genuinely supportive of researchers’ daily workflows. The client is already considering expanding the collaboration.

„Thanks to the new system, data analysis is now much smoother – not only do we save time, but we also gain confidence in the consistency and reliability of the results.”
– ELI Beamlines Team

Got a similar concern? Set up a free consultation.

Book a meeting

Do your machines meet the latest safety standards?
If not, you’re risking more than just failed audits – there’s also a real threat to your employees and your production line.

Modernizing machines in accordance with the Machinery Directive 2006/42/EC is a key step toward improving safety, reliability, and compliance with international standards in manufacturing facilities. Our team includes experienced specialists who have carried out such upgrades in the automotive, food, and pharmaceutical industries. Thanks to their knowledge and hands-on expertise, we are fully prepared to deliver proven, effective solutions tailored to your company’s needs.

Meeting the requirements of Directive 2006/42/EC and the applicable safety levels is not just about legal compliance – it’s also a response to growing expectations from customers and business partners when it comes to adopting the latest safety standards.

Upgrading your machines to comply with Directive 2006/42/EC will bring:

1. Full compliance with legal regulations and international standards,
2. Reduced risk of accidents and better operator protection (safety guards, light curtains, emergency stop switches),
3. Increased machine reliability, fewer breakdowns and downtimes, higher production efficiency,
4. Positive results in safety audits and increased trust from business partners.

All of this leads to greater operational stability and lower maintenance costs. The benefits are measurable – improved safety, production continuity, and a stronger company image as a market-compliant manufacturer.

Modernization eliminates non-compliance with current technical and legal standards. It removes ineffective safety measures in hazardous zones and replaces outdated control systems. It solves the problem of missing as-built documentation and enables full integration of safety systems with PLCs.

It also improves error diagnostics, and the use of technical safety measures – such as guards, light curtains, and emergency stop switches – ensures rapid threat detection and automatic machine shutdown. This significantly reduces operator errors and troubleshooting time. It also helps you pass external audits and implement maintenance procedures aligned with modern standards.

We provide comprehensive machine modernization.
We prepare control system documentation in EPLAN, considering all safety requirements. We carry out mechanical and electrical installation work, update PLC and safety controller software (e.g. PILZ, SICK), and integrate new safety systems with your existing control infrastructure.

Once the modernization is complete, we commission the installation and perform thorough testing to ensure full functionality of all safety systems. Thanks to the experience of our experts, we’re ready to implement these solutions in full compliance with Directive 2006/42/EC and the required safety levels.

Modernizing machines according to Directive 2006/42/EC is a real opportunity to improve workplace safety, regulatory compliance, and production efficiency.
Our team includes specialists with practical experience in the automotive, food, and pharmaceutical sectors. We have the technical capacity and human resources to deliver and implement these solutions in your facility.

Want to bring your machines up to the latest safety standards?

Contact us today – we’ll carry out a free needs assessment and present concrete solutions tailored to your operations!

Book a meeting

Maintaining high cleanliness standards in food production requires modern and reliable solutions. In response to growing regulatory demands and operational needs, one of our specialists carried out a project to modernize the CIP (Clean In Place) station in a production facility. The goal was to transition from a manual, time-consuming process to a fully automated system that meets the latest legal and technological requirements. By implementing modern solutions, we not only streamlined operations but also enabled precise parameter control, process documentation, and real cost savings in daily plant operations. The results of the implementation demonstrated the vast potential of automation—both in terms of efficiency and compliance with industry standards.

The modernization of the CIP cleaning station delivers measurable results from the very first days after implementation. Process automation not only shortens cleaning time but also improves quality, consistency, and overall process reliability. The production facility gains greater control over the cleaning process and the ability to quickly respond to any irregularities.

1. 60% reduction in cleaning time,
2. Lower consumption of water, energy, and detergents,
3. Full compliance with legal regulations,
4. Automatic reporting of each cycle,
5. Improved repeatability and cleaning quality.

The optimized system enables precise control and monitoring of process parameters, as well as real-time tracking of critical values such as chemical concentration and rinsing time.

CIP process automation eliminates many common issues associated with manual or partially automated cleaning. It significantly reduces the risk of incomplete cleaning due to human error and inaccurate chemical dosing. The system ensures full control over key parameters—rinsing time, temperature, and detergent concentration—greatly enhancing the consistency and effectiveness of the cleaning process. With automatic reporting, the need for manual cycle documentation is eliminated. Additionally, access control and user authentication allow for tracking and identifying personnel making changes in the system. Importantly, the system ensures full compliance with EU Regulation 2017/625 of March 15, 2017.

The implementation of an integrated system based on modern industrial automation technologies includes several technical solutions applicable to the CIP cleaning station.

These include:

1. Replacing manually operated valves with automated valves
2. Upgrading the PLC controller to the Siemens S7-1500 model
3. Using Profinet communication modules for improved connectivity
4. Designing the control system in the EPlan environment

Additionally, the system features an HMI panel for entering cleaning recipes, visualization, and process control, along with an RFID reader for user authentication. The system also supports automatic report generation for each cleaning cycle.

Modernizing the CIP cleaning station with advanced industrial automation solutions significantly enhances hygiene, control, and efficiency in production processes. The implementation of cutting-edge technologies ensures full control over every stage of the cleaning process, resulting in improved consistency and compliance with legal regulations. Integrating the system with reporting and user authentication functions increases transparency and operational oversight. The findings from this project highlight that CIP station automation is a real solution to the challenges faced by production plants.

Boost Efficiency and Hygiene Standards in Your Facility! Don’t wait until outdated processes impact your product quality and production costs. Choose modern automation and gain greater control, cost savings, and full regulatory compliance. Contact us today—together, we’ll find the best solution for your production needs!

Got a similar concern? Set up a free consultation.

Book a meeting

Automated utility monitoring is a powerful tool for cost reduction, increased efficiency, and regulatory compliance. Implementing smart meters, a SCADA system, and integrated energy management systems (EMS) in a production facility enables rapid identification of losses, consumption optimization, and reduced downtime risks. 

1. 15% reduction in energy costs achieved through loss elimination and optimized use of electricity, water, and gas. 
2. Increased energy efficiency through optimized equipment operation.
3. Rapid failure and anomaly detection, lowering the risk of downtime by 10%. Full compliance with CO₂ emissions monitoring and utility consumption regulations. 
4. Enhanced production planning with real-time data analysis. 
5. Improved transparency and control over energy costs. 

• Lack of control over actual utility consumption. 
• Delays in detecting failures and leaks.
• Challenges in CO₂ emissions reporting in compliance with regulations.
• Inefficient energy cost management.

Smart meters and sensors monitor consumption in real time, while the SCADA system automatically analyzes data, generates reports, and sends alerts in case of anomalies. Integration with the EMS system enables automatic adjustments to equipment operation, eliminating losses and improving production stability. 

Don’t let inefficient utility consumption drive unnecessary costs. Contact us to learn how modern solutions can benefit your company! 

Got a similar concern? Set up a free consultation.

Book a meeting

PyTango is a Python library that enables communication with the Tango Controls system, a tool used to develop distributed control systems in scientific laboratories such as ESRF, SKAO, MAX IV, and ELETTRA. A bug was discovered in PyTango that remained unresolved for eight years, with repeated attempts to fix it ending in failure. Mateusz Celary from S2Innovation managed to resolve it in just three days—thanks to in-depth code analysis and an unconventional approach: dynamically overriding a function pointer in Boost::Python. This is a story of determination, code analysis, and overcoming organizational constraints.

The main challenge was fixing a bug in PyTango that had remained unresolved for years despite numerous attempts to eliminate it. The issue was related to memory management at the interface between C++ and Python, making it difficult to control object lifecycles.

The PyTango developers had no effective way to solve the problem, and previous approaches had failed. An additional challenge was the time pressure—there were only three days to analyze the issue and determine whether a fix was even possible.

We focused on in-depth code analysis and intensive debugging using GDB. The key was to fully understand the memory management mechanism between the C++ and Python layers, which allowed us to pinpoint the root cause of the bug.

When standard methods failed, we made a bold but effective decision—we applied dynamic function pointer overriding in Boost::Python. This allowed us to control class destructors from Python, enabling more efficient object lifecycle management and eliminating memory-related bugs.

Technologies used: Python, C++, Boost::Python, GDB. However, the key factors in solving the issue were analytical thinking and a creative approach, which enabled precise diagnostics and an effective fix.

Our solution restored system stability and enabled proper memory management between Python and C++. We fixed a critical command that other programs depended on, and in the process, deepened our understanding of PyTango’s inner workings.

Additionally, the implementation of automated tests further increased system reliability and established a solid foundation for future improvements.

The PyTango story shows that even long-standing unresolved bugs can be eliminated through creativity and flexibility in tackling technical challenges. Innovative memory management not only solved the problem but also opened up new opportunities for system optimization.

Do you have an unresolved issue in your code? Our unconventional approach could help you too. Contact us!

Got a similar concern? Set up a free consultation.

Book a meeting

The European Spallation Source (ESS) is a multidisciplinary research facility that houses the world’s most powerful neutron source. The facility employs cryogenic cooling systems to maintain the ultra-low temperatures required for the operation of the linear accelerator. As operations expanded, the need for improved supervision and reduced manual intervention became evident. To address this, the Automated Control Sequence (ACS) was implemented, streamlining processes and enhancing both precision and efficiency.

By automating repetitive processes, ACS significantly streamlined the handling of cryogenic cooling operations and improved supervision at every stage. Operators can now focus more on data analysis and system optimization rather than routine equipment control. Automation increased operational precision, ensuring stable cooling parameters and eliminating inconsistencies.

Thanks to an intuitive graphical user interface (GUI), users have complete control over the process, and real-time data visualization facilitates quick decision-making. The system is designed for future expansion—the next planned update includes replacing Excel as the database to enhance efficiency.

Previously, each device required a dedicated operator. With 27 cryomodules being cooled simultaneously, this created significant organizational and technical challenges. Manual operation of so many modules would have required multiple operators or placed excessive strain on a single person, leading to delays and inconsistencies.

The lack of centralized coordination made synchronization difficult, as operators had to repeatedly perform the same tasks, wasting time on routine operations. With the increasing number of devices, the risk of human error grew, and manual control did not allow for precise parameter adjustments. Automating the cooling process was essential to improve operational consistency and reduce excessive personnel involvement.

The Automated Control Sequence (ACS) was developed to fully automate cryomodule control, eliminating the need for manual operation. Each module was equipped with a dedicated PLC controller, while a master unit coordinates their operation, ensuring process synchronization. Integration with EPICS allows real-time monitoring of parameters and instant responses to changing conditions.

To execute the project:

• Siemens PLC controllers were programmed using TIA Portal.
• Python scripts were used to automatically generate PLC code, reducing update time significantly.
• The Phoebus environment was used to develop the user interface, providing operators with easy access to system status information.
• The interface also enabled intuitive visualizations, supporting operational decision-making.

Initially, Excel was used as the database for storing operational sequences. However, due to the growing project scale, it is planned to be replaced with a more efficient data management system for better future modifications and ACS optimization.

The implementation of ACS at ESS optimized cryogenic cooling control, improving repeatability and reliability. Automation reduced routine tasks performed by operators, minimizing errors and enhancing system consistency. Future improvements include integrating a more efficient database, which will further streamline information management and system development.

Thinking about process automation? Start by analyzing your current procedures and choose a flexible control system that grows with your business. Implement solutions that speed up operations and boost efficiency. Contact us to learn how we can help automate your process!

Got a similar concern? Set up a free consultation.

Book a meeting