By working with communities beyond high-energy physics, we are able to ensure maximum relevancy for CERN openlab’s work, as well as learning and sharing both tools and best practices across scientific fields. Today, more and more research fields, such as medical research or space and Earth observation research, are driven by large quantities of data, and thus experience ICT challenges comparable to those at CERN. CERN openlab’s mission rests on three pillars: technological investigation, education, and dissemination. Collaborating with research communities and laboratories outside the high-energy physics community brings together all these aspects. Challenges related to the life sciences, medicine, astrophysics, and urban/environmental planning are all covered in this section, as well as scientific platforms designed to foster open collaboration.

 

SmartANOMALY

Project goal

The SmartANOMALY project is an evolution and broadening of the SmartLINAC project, which launched in June 2019. The main goal of the original project was to create a platform for anomaly detection and maintenance planning for linear accelerators, which are used widely in medicine and high-energy physics research.

R&D topic
Applications in other disciplines
Project coordinator(s)
Alberto Di Meglio
Team members
Yann Donon

Collaborators

Project background

Technologies related to artificial intelligence (AI) are opening up new possibilities for anomaly detection. Given the array of large particle accelerators at CERN, the Organization has significant expertise in detecting anomalies in highly complex systems. This expertise has the potential to be applied to a range of scientific and industrial activities including (but not limited to) other fields where particle accelerators are used, such as medicine.

This project is being carried out in the context of CERN's strategy for knowledge transfer to medical applications, led by CERN's Knowledge Transfer group

Recent progress

After a year of development, promising results were achieved, demonstrating the potential of our innovative algorithms for detecting anomalies — as well as perhaps even predicting their effects to some extent. Today, the project’s primary focus is on medical accelerators. However, we see potential in training our solution on more sources, such as on compressor engines or complex industrial processes.

It is common practice to use alternative data sets when training anomaly-detection systems. Therefore, the distinguishing aspect of our research is that several approaches, based on statistics and neural-network technologies, are being combined in order to offer a system that can be adapted to different sources.

Given that demand for such tools is growing rapidly, we believe the time is right to formally enlarge the scope of the research started through SmartLINAC. Thus, we have created our new, broader SmartANOMALY project. 

Next steps

The SmartANOMALY project is an evolution and broadening of the SmartLINAC project, which launched in June 2019. The main goal of the original project was to create a platform for anomaly detection and maintenance planning for linear accelerators, which are used widely in medicine and high-energy physics research.

Publications

    Y. Donon, Smart Anomaly Detection and Maintenance Planning Platform for Linear Accelerators (3 October). Presented at the 27th International Symposium Nuclear Electronics and Computing (NEC’2019), Montenegro, 2019. cern.ch/go/nb9z

Presentations

    Y. Donon, Smart Anomaly Detection and Maintenance Planning Platform for Linear Accelerators (3 October). Presented at the 27th International Symposium Nuclear Electronics and Computing (NEC’2019), Montenegro, 2019.
    Y. Donon, Anomaly detection in noised time series: the challenge of CERN’s LINAC4 (24 January). Presented at The Open Data science meetup #3, Samara, 2020. cern.ch/go/9PZD

CERN Living Lab 

Project goal

The project goal is to develop a big-data analytics platform for large-scale studies of data under special constraints, such as information that is privacy-sensitive, or that has a varying level of quality, associated provenance information, or signal-to-noise ratio. Ethical considerations are also considered when necessary. This will serve as a proof-of-concept for federating and analysing heterogeneous data from diverse sources, in particular for medical and biological research, using ideas and expertise coming from CERN and the broader high-energy physics community.

R&D topic
Applications in other disciplines
Project coordinator(s)
Alberto Di Meglio
Team members
Taghi Aliyev, Jose Cabrero, Anna Ferrari
Collaborator liaison(s)
David Manset (be-studys), Marco Manca (SCImPULSE)

Collaborators

Project background

CERN is a living laboratory, with around 15,000 people coming to work at its main campuses every day. For operational purposes, CERN collects data related to health, safety, the environment, and other aspects of daily life at the lab. Creating a platform to collate and enable intelligent management and use of this data — while respecting privacy and other ethical and legal obligations — offers the potential to improve life at the lab. At the same time, such a platform provides an ideal testbed for exploring new data analytics technologies, algorithms and tools, including ML/DL methods, encryption schemes, or block-chain-based ledgers. It also provides a natural bridge to collaborate with other scientific research domains, such as medical research and biology.

This project is being carried out in the context of CERN's strategy for knowledge transfer to medical applications, led by CERN's Knowledge Transfer group.

Recent progress

The CERN Living Lab project started formally in June 2019. A kick-off meeting was held with all project partners to discuss in detail shared interests and objectives. A second meeting took place in December 2019, focusing on the architecture and requirements for the big-data analytics platform. The platform architecture was defined and agreed. Also, four specific sub-projects were defined to address the data life-cycle in the presence of sensitive information, the data ingestion from foreign sources, the possibility of dynamically detecting and addressing the level of privacy protection required by different data transfer requests, and the use of homomorphic encryption as a possible privacy-preserving approach for cloud-based data analysis

Next steps

In 2020, a proof-of-concept platform will be established, and a number of selected use cases will be deployed. Specifically, investigations will include classification and detection of the symptoms of Parkinson’s disease from wearable devices, optimisation of homomorphic encryption techniques for deep learning, and medical image analysis.


Presentations

    T. Aliyev, Meaningful Control of AI and Machine Ethics (7 June). Presented at Big Data in Medicine: Challenges and Opportunities, CERN, Geneva, 2019. cern.ch/go/J7CF
    A. Di Meglio, The CERN Living Lab Initiative (20 June). Presented at CERN Information Technology for the Hospitals, HUG, Geneva, 2019. cern.ch/go/Fld8
    T. Aliyev, Interpretability and Accountability as Necessary Pieces for Machine Ethics (2 July). Presented at Implementing Machine Ethics Workshop, UCD, Dublin, 2019. cern.ch/go/7c6d

Smart platforms for science

Project goal

The project goal is to design a platform that can analyse data collected from user interactions and can process this information in order to provide recommendations and other insights, thus helping improve the performance and relevance of user searches or learning objectives.

R&D topic
Applications in other disciplines
Project coordinator(s)
Alberto Di Meglio
Team members
Taghi Aliyev
Collaborator liaison(s)
Marco Manca (SCImPULSE), Mario Falchi (King’s College London)

Collaborators

Project background

Data-analysis systems often collect and process very different types of data. This includes not only the information explicitly entered by users (“I’m looking for…”), but also metadata about how the user interacts with the system and how their behaviour changes over time based on the results they get. Using techniques such as natural language processing (NLP) and smart chatbots, it is possible to achieve improved interaction between humans and machines, potentially providing personalised insights based on both general population trends and individual requests. Such a system would then be able to recommend further searches, actions, or links that may have not occurred to the user.

Such an approach could, for example, be used to design better self-help systems, automated first-level medical services, more contextual and objective-aware search results, or educational platforms that are able to suggest learning paths that address specific student needs.

This project is being carried out in the context of CERN's strategy for knowledge transfer to medical applications, led by CERN's Knowledge Transfer group.

Recent progress

The concept of the Smart Platforms project emerged in 2019 as a spin-off of the application of NLP techniques to genomic analysis in the GeneROOT project.

In 2019, a few initial discussions about possible applications were started in collaboration with educational institutes and public administrations, with the goal of developing the concept of smart chatbots that are able to improve human-machine interaction. As the project moved into the proof-of-concept phase, it became clear that the need to understand issues related to data-privacy and information sharing are still a critical roadblock for systems like this. The project was therefore merged into the CERN Living Lab, through which such concerns can be better addressed.

Next steps

The project has been merged into the CERN Living Lab as part of a general initiative to understand the implications of processing personal data and the related ethical constraints.

Publications

    A. Manafli, T. Aliyev: Natural Language Processing for Science. Information Retrieval and Question Answering. Summer Student Report, 2018. cern.ch/go/Z9l9

Presentations

    A. Di Meglio, Introduction to Multi-disciplinary Platforms for Science, (24 January). Presented at CERN openlab Technical Workshop, CERN, Geneva, 2019. cern.ch/go/XNt9
    T. Aliyev, Smart Data Analytics Platform for Science (1 November). Presented at i2b2 tranSMART Academic Users Group Meeting, Geneva, 2018.
    T. Aliyev, AI in Science and Healthcare: Known Unknowns and potential in Azerbaijan (December). Presented at Bakutel Azerbaijan Tech Talks Session, Baku, 2018.

Humanitarian AI applications for satellite imagery

Project goal

This project is making use of expertise in artificial intelligence (AI) technologies at CERN to support a UN agency. Specifically, we are working on AI approaches to help improve object recognition in the satellite imagery created to support humanitarian interventions. Such satellite imagery plays a vital role in helping humanitarian organisations plan and coordinate responses to natural disasters, population migrations, and conflicts.

R&D topic
Applications in other disciplines
Project coordinator(s)
Sofia Vallecorsa, Federico Carminati
Team members
Taghi Aliyev, Yoann Boget
Collaborator liaison(s)
Lars Bromley

Collaborators

Project background

Since 2002, CERN has hosted UNOSAT, the Operational Satellite Applications Programme of UNITAR (The United Nations Institute for Training and Research) on the laboratory’s premises. UNOSAT acquires and processes satellite data to produce and deliver information, analysis, and observations to be used by the UN or national entities for emergency response, to assess the impact of a disaster or a conflict, or to plan sustainable development in the face of climate change.

At the heart of this project lies the idea of developing machine-learning techniques that can help speed up analysis of satellite imagery. For example, predicting and understanding the movement of displaced persons by identifying refugee shelters can be a long, labour-intensive task. This project is working to develop machine-learning techniques that could greatly reduce the amount of time needed to complete such tasks.

Recent progress

Refugee camps often consist of more than 10,000 shelters and may need to be re-analysed several times in order to understand their evolution. Manual analysis typically leads to very high-quality output, but is very time-consuming. We have therefore worked with region-based convolutional neural networks to improve detection of new shelters in refugee camps, taking into account prior knowledge regarding the positions of existing shelters. The results were promising and the data pipeline created by our summer students has now been adapted and put to use by the UNOSAT experts. The retrained model yielded an average precision/recall score of roughly 80% and reduced the time needed for the task by a factor of 200 in some areas.

More recently, we also addressed the challenge of simulating synthetic high-resolution satellite images. High-resolution satellite imagery is often licensed in such a way that makes it difficult to share it across UN partners and academic organisations. This reduces the amount of image data available to train deep-learning models, thus hampering research in this area. We have developed a generative adversarial network (GAN) that is capable of generating realistic satellite images of refugee camps images. Our tool was initially based on a progressive GAN approach developed by NVIDIA. We have now developed this further, such that it can combine multiple simulated images into a cohesive larger image of roughly 5 million pixels.

Several other lines of investigation — all related to AI technologies — are also being pursued within the scope of this project.

Next steps

Next year, we will pursue the initial work carried out on the GAN model in 2019 in a number of different directions. We will carry out a detailed performance study and will implement a distributed approach for parallel network training, as well optimising the use of computing resources. This should help us to reduce training time for the model and increase the maximum image size.

Publications

    N. Lacroix, T. Aliyev, L. Bromley: Automated Shelter Recognition in Refugee Camps. CERN openlab Summer Student Report. Published on ZENODO, 2019. cern.ch/go/v6rn

Presentations

    Y. Boget, ProGAN on Satellite images (15 August). Presented at CERN openlab summer student lightning talk session, Geneva, 2019. cern.ch/go/P6NV
    Y. Boget, S. Vallecorsa, Deep Learning for Satellite Imagery (24 September). Presented at IXPUG Annual Conference, Geneva, 2019. cern.ch/go/m9n6

BioDynaMo

Project goal

We are aiming to create a platform through which life scientists can easily create, run and visualise three-dimensional biological simulations. Built on top of the latest computing technologies, the BioDynaMo platform will enable users to perform simulations of previously unachievable scale and complexity, making it possible to tackle challenging scientific research questions.

 

R&D topic
Applications in other disciplines
Project coordinator(s)
Roman Bauer, Fons Rademakers
Team members
Lukas Breitwieser, Ahmad Hesam
Collaborator liaison(s)
Uri Nevo, Marco Durante, Vasilis Vavourakis

Collaborators

Project background

Within the life-sciences community, computer simulation is being used more and more to model increasingly complex biological systems. Although many specialised software tools exist, establishing a high-performance, general-purpose platform would be a major step forward. CERN is therefore contributing its deep knowledge in large-scale computing to this collaboration with Newcastle University in the UK and other institutions, supported by Intel. Together, we are working to develop a unique platform. This project is cofinanced by the CERN budget for knowledge transfer to medical applications.

ImmunoBrain Checkpoint, GSI Darmstadt and the University of Cyprus are also collaborating in BioDynaMo. Find out more about the project on the collaboration page: biodynamo.org/.

Recent progress

We further integrated core CERN technologies into BioDynaMo. ROOT Notebooks provide life scientists with a web-based interface for (i) creating, running and visualising simulations, (ii) performing powerful analysis, and (iii) working interactively with a graphical user interface. Furthermore, users are now able to explore a large model parameter space on distributed computing infrastructures (e.g. cloud computers and computer clusters).

Based on feedback from our users, we improved BioDynaMo’s API. Now, life-scientists can translate an idea into a simulation faster than before. We were also able to connect BioDynaMo with another simulator, to benefit from the strengths of agent-based and continuum-based modelling for cancer research.

In addition, we succeeded in integrating systems biology markup language (SBML) into BioDynaMo. This makes it possible for life scientists to reuse the large database of existing SBML models to simulate chemical reaction networks within each BioDynaMo simulation object.

Finally, BioDynaMo was featured in CERN's official teachers and students programme in 2019. Teachers and students from Dutch high schools learned through hands-on sessions how ICT technologies developed at CERN help in tackling challenges in the biomedical fields.

Next steps

BioDynaMo is currently able to simulate millions of cells on one server. To improve the performance yet further, we will focus on two main aspects. First, we will continue development on the distributed runtime, to combine the computational resources of many servers. Second, we will improve hardware acceleration to fully utilise (multiple) GPUs in a system. This will not only reduce runtime on high-end systems, but will also benefit users that work on a standard desktop or laptop.

Finally, thanks to new projects funded by the UK Engineering and Physical Sciences Research Council and the UK Medical Research Council in 2019, BioDynaMo has been extended for applications related to cryopreservation and the representation of realistic neural networks. Work in these areas will continue in 2020.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Publications

    L. Breitwieser, BioDynaMo: A New Platform for Large-Scale Biological Simulation (Master’s thesis), Graz University of Technology, Austria, 2016. cern.ch/go/z67t
    L. Breitwieser, R. Bauer, A. Di Meglio, L. Johard, M. Kaiser, M. Manca, M. Mazzara, F. Rademakers, M. Talanov, The BioDynaMo project: Creating a platform for large-scale reproducible biological simulations. CEUR Workshop Proceedings, 2016. cern.ch/go/Xv8l
    R. Bauer, L. Breitwieser, A. Di Meglio, L. Johard, M. Kaiser, M. Manca, M. Mazzara, F. Rademakers, M. Talanov, A. D. Tchitchigin, The BioDynaMo project: experience report. In Advanced Research on Biologically Inspired Cognitive Architectures (pp. 117-125). IGI Global, 2017. cern.ch/go/dp77
    A. Hesam, Faster than the Speed of Life: Accelerating Developmental Biology Simulations with GPUs and FPGAs (Master’s thesis), Delft University of Technology, Netherlands, 2018. cern.ch/go/f9v6
    J. de Montigny, A. Iosif, L. Breitwieser, M. Manca, R. Bauer, V. Vavourakis, An in silico hybrid continuum-/agent-based procedure to modelling cancer development: interrogating the interplay amongst glioma invasion, vascularity and necrosis. Methods, 2020. cern.ch/go/6vrm

Presentations

    K. Kanellis, Scaling a biological simulation platform to the cloud (15 August), Presented at CERN openlab summer students’ lightning talks, Geneva, 2017. cern.ch/go/d9nV
    L. Breitwieser, BioDynaMo (21 September), Presented at CERN openlab Open Day, Geneva, 2017. cern.ch/go/lNP9
    L. Breitwieser & A. Hesam, BioDynaMo: Biological simulation in the cloud (1 December), Presented at CERN IT technical forum, Geneva, 2017. cern.ch/go/m9Kw
    A. Hesam, Biodynamo project status and plans (11 January). Presented at CERN openlab Technical Workshop, Geneva, 2018. cern.ch/go/F8Cl
    L. Breitwieser, BioDynaMo (1 February). Presented at University Hospital of Geneva Café de l'Innovation, Geneva, 2018.
    L. Breitwieser, The Anticipated Challenges of Running Biomedical Simulations in the Cloud (12 February). Presented at Early Career Researchers in Medical Applications @ CERN, Geneva, 2018. cern.ch/go/spc8
    N. Nguyen, Distributed BioDynaMo (16 August). Presented at CERN openlab summer students' lightning talks, Geneva, 2018.
    A. Hesam, Faster than the Speed of Life: Accelerating Developmental Biology Simulations with GPUs and FPGAs (31 August). Master’s thesis defense, Delft, 2018. cern.ch/go/f9v6
    L. Breitwieser, The BioDynaMo Project: towards a platform for large-scale biological simulation (17 September). Presented at DIANA meeting, Geneva, 2018. cern.ch/go/kJv7
    L. Breitwieser, BioDynaMo (23 January). Presented at CERN openlab workshop, Geneva, 2019. cern.ch/go/7RlF
    R. Bauer, Computational modelling and simulation of biophysical dynamics in medicine (7 June). Presented at Big Data in Medicine: Challenges and Opportunities, Geneva, 2019. cern.ch/go/xf9F
    F. Rademakers, BioDynaMo (20 June). Presented at Hôpitaux Universitaires de Genève, Geneva, 2019.
    J. L. Jennings, Computational Modelling of cryopreservation using the BioDynaMo software package CryoDynaMo (22 July). Presented at the Society for Cryobiology Conference, San Diego, 2019.
    J. de Montigny, Computational modelling of retinal ganglion cell development (July). Presented at UK Neural Computation, Nottingham, 2019.
    G. De Toni, Improvements on BioDynaMo Build System (13 August). Presented at the CERN openlab summer student lightning talk session, Geneva, 2019. cern.ch/go/xt68
    G. De Toni, Improvements on BioDynaMo Build System (13 August). Presented at the CERN openlab summer student lightning talk session, Geneva, 2019. cern.ch/go/xt68
    L. Breitwieser, BioDynaMo Project Update (26 September). Presented at CERN Medical Application Project Forum, Geneva, 2019.
    A. Hesam, Simulation Master Class (26 September). Presented at CERN's Dutch Language Teachers Programme, Geneva, 2019. cern.ch/go/hKL7
    A. Hesam, Simulation Master Class (11 October). Presented at CERN's Dutch Language Students Programme (NVV Profielwerkstukreis), Geneva, 2019.
    L. Breitwieser, A. Hesam, The BioDynaMo Project (21 October). Presented at EmLife Meeting, Geneva, 2019.
    L. Breitwieser, The BioDynaMo Software Part I (2 December). Presented at the BioDynaMo Collaboration Meeting, Zurich, 2019.
    A. Hesam, The BioDynaMo Software Part II (2 December). Presented at the BioDynaMo Collaboration Meeting, Zurich, 2019.
    R. Bauer, BioDynaMo: A platform for computational models and simulations of biological systems, Presented at CERN Knowledge Exchange Event, Daresbury, 2019.