Generative AI for model driven engineering

Generative AI and large language models (LLMs), such as Copilot and ChatGPT can complete code based on initial fragments written by a developer. They are integrated in software development environments such as VS code. Many papers analyse the advantages and limitations of these approaches for code generation. Besides some deficiencies, the produced code is often correct and the results are improving.

However, a surprisingly small amount of work has been done in the context of software modeling. The paper from Cámara et al. concludes that while the performance of the current LLMs for software modeling is still limited (in contrast to code generation), there is a need that (in contrast to code generation) we should adapt our model-based engineering practices to these new assistants and integrate these into MBSE methods and tools.

The goal of this post-doc is to explore generative AI in the context of system modeling and associated tool support. For instance, AI assistance can support the completion, re-factoring and analysis (for instance identified design patterns or anti-patterns) at the model level. Propositions are discussed in the team and in a second step prototyped and evaluated the mechanism in the context of the open-source UML modeler Papyrus.

Development of noise-based artifical intellgence approaches

Current approaches to AI are largely based on extensive vector-matrix multiplication. In this postdoctoral project we would like to pose the question, what comes next? Specifically we would like to study whether (stochastic) noise could be the computational primitive that the a new generation of AI is built upon. This question will be answered in two steps. First, we will explore theories regarding the computational role of microscopic and system-level noise in neuroscience as well as how noise is increasingly leveraged in machine leaning and artificial intelligence. We aim to establish concrete links between these two fields and, in particular, we will explore the relationship between noise and uncertainty quantification.
Building on this, the postdoctoral researcher will then develop new models that leverage noise to carry out cognitive tasks, of which uncertainty is an intrinsic component. This will not only serve as an AI approach, but should also serve as a computational tool to study cognition in humans and also as a model for specific brain areas known to participate in different aspects of cognition, from perception to learning to decision making and uncertainty quantification.
Perspectives of the postdoctoral project should inform how future fMRI imaging and invasive and non-invasive electrophysiological recordings may be used to test theories of this model. Additionally, the candidate will be expected to interact with other activates in the CEA related to the development of noise-based analogue AI accelerators.

Eco-innovation of insulating materials by AI, for the design of a future cable that is long-lasting, resilient, bio-sourced and recyclable.

This topic is part of a larger upcoming project for the AI-powered creation of a new electrical cable for future nuclear power plants. The goal is to design cables with a much longer lifetime than existing cables in an eco-innovative approach.
The focus is on the cable insulation because it is the most critical component for the application and the most sensitive to aging. The current solution is based on adding additives (anti-rad and antioxidants) to the insulation to limit the effects of irradiation and delay aging as much as possible. However, there is another solution that has never been tested before: self-repairing materials.
The project to which this topic is attached aims to design and manufacture several test model of insulation specimens. With several test characterization protocols, in order to verify the gain in terms of reliability and resilience. The results obtained will begin to fill a future database for the AI platform Expressif, developed at CEA List, which will be used to design the future cable.

Co-design strategy (SW/HW) to enable a structured spatio-temporal sparsity for NN inference/learning

The goal of the project is to identify, analyze and evaluate mechanisms for modulating the spatio-temporal sparsity of activation functions in order to minimize the computational load of transformer NN model (learning/inference). A combined approach with extreme quantization will also be considered.
The aim is to jointly refine an innovative strategy to assess the impacts and potential gains of these mechanisms on the model execution under hardware constraints. In particular, this co-design should also enable to qualify and to exploit a bidirectional feedback loop between a targeted neural network and a hardware instantiation to achieve the best tradeoff (compactness/latency).

LLMs hybridation for requirements engineering

Developing physical or digital systems is a complex process involving both technical and human challenges. The first step is to give shape to ideas by drafting specifications for the system to be. Usually written in natural language by business analysts, these documents are the cornerstones that bind all stakeholders together for the duration of the project, making it easier to share and understand what needs to be done. Requirements engineering proposes various techniques (reviews, modeling, formalization, etc.) to regulate this process and improve the quality (consistency, completeness, etc.) of the produced requirements, with the aim of detecting and correcting defects even before the system is implemented.
In the field of requirements engineering, the recent arrival of very large model neural networks (LLMs) has the potential to be a "game changer" [4]. We propose to support the work of the functional analyst with a tool that facilitates and makes reliable the writing of the requirements corpus. The tool will make use of a conversational agent of the transformer/LLM type (such as ChatGPT or Lama) combined with rigorous analysis and assistance methods. It will propose options for rewriting requirements in a format compatible with INCOSE or EARS standards, analyze the results produced by the LLM, and provide a requirements quality audit.

Instrumental development on the Jeol NeoARM TEM for rapid and low-dose mapping of EELS and strain fields

The primary objective of our Post-Doctoral Researcher position is to focus on instrumental development for the NeoARM 200F. Collaborating closely with a research software engineer already familiar with TEM automation, the team will be tasked with the automation of alignment, interface, and data stream processing for the filter and various detectors to enable cutting-edge and valuable applications. The first part of the project will focus on the rapid, precise, low-dose mapping of material strain at the nanometer scale. To achieve this, we are utilizing Nanomegas ASTAR along with other universal scan generators to acquire scanning precession electron diffraction (SPED) data. CEA has a well-established expertise in strain mapping in TEM [e.g., D. Cooper et al., Micron 80 (2016) 145]. Currently, there is a strong desire to advance the boundaries of this technique to generate real-time, high-quality results. Following the complete integration of the NeoARM for strain analysis, we will assess the advantages provided by components such as the TimePix3 detector and energy filtering for diffraction patterns. Simultaneously, the techniques developed for the SPED datasets will be extended other applications such as orientation/phase or electric field mapping. The second part of the project will focus on the TimePix3 detector for electron energy-loss spectroscopy (EELS). In frame-based acquisition mode, the performance of STEM-EELS using the TimePix3 detector will be assessed and highlighted (e.g., simultaneous EELS/EDX tomography). New strategies to enhance the EELS signal-to-noise ratio for direct electron cameras will also be explored. The exploitation of the TimePix3 in event-based mode, including EELS/EDX co-incidence and low-dose, high-speed imaging will also enter into the project as appropriate.

Development of a new spectrometer for the characterization of the radionuclide-based neutron sources

Since few years, the LNHB is developing a new instrument dedicated to the neutron spectrometry, called AQUASPEC. The experimental device consists of a polyethylene container that is equipped with a central channel accommodating the source and 12-measurement channels (in a spiral formation) around the source, into which detectors can be placed. The container is filled with water in order to moderate neutrons emitted from the source. Measurements have performed with 6Li-doped plastic scintillators, optimized for the simultaneous detection of fast neutrons, thermal neutrons and gamma rays through the signal processing based on pulse shape discrimination (PSD). The spectrum reconstruction is performed with an iterative ML-EM or MAP-EM algorithm, by unfolding experimental data through the detector's responses matrix calculated with MCNP6 code. The candidate will work in the general way on issues related to the neutron spectrometry in the laboratory: Contribution to the development and validation of the new spectrometer AQUASPEC; Participation to the sources measurements and working on aspects of neutron detection and signal processing, in particular issue of the discrimination of neutron/gamma based on the pulse shape discrimination technique (PSD); Usage of Monte Carlo simulation codes and algorithms to reconstruct initial neutron energy distribution; Investigation and integration of information related to neutron/gamma coincidence specific to the XBe type sources.

Contribution to the metrological traceability of emerging alpha-emitting radiopharmaceuticals in the framework of the european AlphaMet project (Metrology for Emerging Targeted Alpha Therapies)

The Laboratoire national Henri Becquerel (LNE-LNHB) at CEA/Saclay is the laboratory responsible for the french references in the field of ionizing radiation. The LNHB is involved in the european EPM AlphaMet (Metrology for Emerging Targeted Alpha Therapies) submitted under the Metrology support for Health call (2022) to provide metrological support for clinical and preclinical studies; it began in September 2023 for a total duration of three years. The project comprises four Work Packages (WP) targeting different issues, with WP1 in particular dedicated to activity metrology and nuclear data measurements for imaging and dosimetry. This project aims at to improve the metrological traceability of emerging alpha-emitting radiopharmaceuticals such as 211At, 212Pb/212Bi, 225Ac.
The candidate will participate in the various tasks defined as part of the European AlphaMet project in which the LNHB is involved. Radiation-matter simulations will be carried out to study the response of the laboratory's ionisation chambers in various situations concerning: (i) the evolution of the response during the in-growth of the ?-emitting progeny of 225Ac, (ii) the quantification of the influence of the 210At impurity in the case of the measurement of 211At, and (iii) the search for a long-lived radionuclide surrogate of 212Pb for the quality control of dose calibrators. The candidate will also be involved in setting up a new device aimed at improving the linearity of the measurement of half-life with an ionization chamber. During the post-doctoral stay at LNHB, the candidate will interact with the various partners in the AlphaMet project (activity metrology laboratories, hospitals, clinical study centres).
The initial duration of the post-doctorate is 12 months (renewable) at the Laboratoire National Henri Becquerel (CEA/Saclay). It is hoped to start in the first half of 2024.

Development of piezoelectric resonators for power conversion

CEA-Leti has been working to improve energy conversion technologies for over 10 years. Our research focuses on designing more efficient and compact converters by leveraging GaN-based transistors, thereby setting new standards in terms of ultra-fast switching and energy loss reduction.
In the pursuit of continuous innovation, we are exploring innovative paths, including the integration of piezoelectric mechanical resonators. These emerging devices, capable of storing energy in the form of mechanical deformations, offer a promising perspective for increased energy density, particularly at high frequencies (>1 MHz). However, the presence of parasitic resonance modes impacts the overall efficiency of the system. Therefore, we are in need of an individual with skills in mechanics, especially in vibrational mechanics, to enhance these cleanroom-manufactured micromechanical resonators.
You will be welcomed in Grenoble within a team of engineers, researchers and doctoral students, dedicated to innovation for energy, which combines the skills of microelectronics and power systems from two CEA institutes, LETI and LITEN, close to the needs of the industry (http://www.leti-cea.fr/cea-tech/leti/Pages/recherche-appliquee/plateformes/electronique-puissance.aspx).
If you are a scientifically inclined mind, eager to tackle complex challenges, passionate about seeking innovative solutions, and ready to contribute at the forefront of technology, this position/project represents a unique opportunity. Join our team to help us push the boundaries of energy conversion.

References : http://scholar.google.fr/citations?hl=fr&user=s3xrrcgAAAAJ&view_op=list_works&sortby=pubdate

Development and application of TERS/TEPL technique for advanced characterization of materials

TERS/TEPL (Tip-Enhanced Raman Spectroscopy and Tip-Enhanced Photoluminescence) are powerful analytical techniques developed for nanoscale material characterization. The recent acquisition of a unique and versatile TERS/TEPL equipment at PFNC (Nano-characterization Platform) of CEA LETI opens up new horizons for materials characterization. This tool combines Raman spectroscopy, photoluminescence, and scanning probe microscopy. It features multi-wavelength capabilities (from UV to NIR), allowing a wide range of applications and providing unparalleled insights into the composition, structure, and mechanical/electrical properties of materials at nanoscale resolution. The current project aims to develop and accelerate the implementation of the TERS/TEPL techniques at PFNC to fully exploit its potential in diverse ongoing projects at CEA-Grenoble (LETI/LITEN/IRIG) and with its partners.

Top