Portail HAL UNIV-RENNES
Not a member yet
180610 research outputs found
Sort by
A safari across France: soil fauna insights from a nationwide soil quality monitoring program
No full textInternational audienceSoil biodiversity is fundamental to ecosystem functioning but remains underrepresented in conservation policies and large-scale monitoring. Here, we present RMQS-Biodiversity, a nationwide soil biodiversity survey integrated into the French Soil Quality Monitoring Network (RMQS), and illustrate its potential for soil ecology research. In this pioneer study, we examine three major ecological aspects: (i) how systematic grid-based sampling captures micro-food web patterns using nematode communities, (ii) the spatial turnover of detritivore communities (Collembola, Isopoda, Diplopoda) in response to environmental and geographic gradients, and (iii) the influence of macroecological drivers on predator (Carabidae) morphological traits. Across 69 sites, we identified a few widespread species coexisting with numerous rare taxa, underscoring the value of large-scale surveys for detecting cryptic biodiversity. Nematode indicators revealed high variability in food web structure across land uses, with increased facultative phytophagous nematodes in forests. Isopods and diplopods were strongly structured by dispersal constraints, while springtails exhibited weaker environmental responses, likely due to their higher dispersal capacity. Additionally, sexual size dimorphism in Carabidae varied by habitat, with female-biased dimorphism in closed habitats but no dimorphism in open environments, highlighting habitat stability's role in shaping morphological traits. This study demonstrates the value of multi-taxon, multi-trophic biodiversity assessments in long-term soil monitoring. RMQS-Biodiversity provides a robust framework for soil biodiversity monitoring and conservation, refining bioindicators of soil quality and informing policies such as the EU Soil Monitoring Law
Dynamic Agent Generation for Self-Adaptive Root Cause Analysis
No full textInternational audienceModern software systems increasingly rely on microservice architectures, which enhance modularity and resilience but produce vast amounts of heterogeneous observability data—logs, metrics, and traces—to ensure reliable operation and early failure detection. Performing Root Cause Analysis (RCA) on such data is challenging due to its scale, heterogeneity, and evolving structure, which hinder effective correlation and reasoning across modalities. Although recent studies have explored statistical techniques, graph-based models, and Large Language Models (LLMs)-based agents for RCA, most remain static and task-specific, lacking the adaptability and coordination required to handle evolving diagnostic contexts. This paper introduces a self-adaptive agent generation framework that leverages LLMs to dynamically compose and orchestrate adapted diagnostic agents at runtime according to each anomaly’s characteristics. Two main agents drive this process: a Parser Agent that interprets natural-language queries and builds structured task specifications, and an Executor Agent that adapts and coordinates adapted agents analyzing multimodal observability data through a shared memory space. Experiments on Nezha (fault-injected) and OpenRCA (real-world) datasets show up to 12% and 22% gains in diagnostic accuracy, confirming the framework’s effectiveness in adaptive reasoning, coordination, and interpretable root cause identification
Knock-Knock: Black-Box, Platform-Agnostic DRAM Address-Mapping Reverse Engineering
No full textInternational audienceModern Systems-on-Chip (SoCs) employ undocumented linear address-scrambling functions to obfuscate DRAM addressing, which complicates DRAM-aware performance optimizations and hinders proactive security analysis of DRAM-based attacks; most notably, Rowhammer. Although previous work tackled the issue of reversing physical-to-DRAM mapping, existing heuristic-based reverse-engineering approaches are partial, costly, and impractical for comprehensive recovery. This paper establishes a rigorous theoretical foundation and provides efficient practical algorithms for black-box, complete physical-to-DRAM address-mapping recovery.We first formulate the reverse-engineering problem within a linear algebraic model over the finite field GF(2). We characterize the timing fingerprints of row-buffer conflicts, proving a relationship between a bank addressing matrix and an empirically constructed matrix of physical addresses. Based on this characterization, we develop an efficient, noise-robust, and fully platform-agnostic algorithm to recover the full bankmask basis in polynomial time, a significant improvement over the exponential search from previous works. We further generalize our model to complex row mappings, introducing new hardware-based hypotheses that enable the automatic recovery of a row basis instead of previous human-guided contributions.Evaluations across embedded and server-class architectures confirm our method's effectiveness, successfully reconstructing known mappings and uncovering previously unknown scrambling functions. Our method provides a 99% recall and accuracy on all tested platforms. Most notably, Knock-Knock runs in under a few minutes, even on systems with more than 500GB of DRAM, showcasing the scalability of our method. Our approach provides an automated, principled pathway to accurate DRAM reverse engineering
Chemical contamination of black soldier fly larvae raised on EU-authorized or unauthorized substrate
No full textInternational audienceBlack soldier fly larvae (BSFL), Hermetia illucens, can efficiently convert biowaste into high-quality protein for the feed industry. However, biowaste can contain several chemical hazards and their fate in BSFL remain largely unexplored, even though they can pose a threat for both the insect sector and the higher levels of the trophic chain. The purpose of this study was to evaluate the chemical contamination of BSFL reared on authorized biowaste (wheat bran, carrots, apricots, salad) or unauthorized biowaste in the EU (school canteen and supermarket biowaste). PFAS were not quantifiable in the substrates in contrast with persistent organic pollutants, pesticides and trace metal elements. The chlormequat pesticide was the only pesticide quantified in BSFL (15 µg/kg), slightly above the maximum residue limit in the EU for feed (10 µg/kg). Concentrations of PCBs and PCDD/Fs in BSFL were below the maximum limit (ML) but bioaccumulation factors up to 5 were obtained. As and Cd were highly bioaccumulated with concentrations approaching the ML for Cd (0.5 mg/kg) or 8 times below for As (ML = 2 mg/kg). The unauthorized substrate samples tested led to higher chemical safety risks in BSFL, with 10 times higher concentrations of As than in BSFL reared on authorized substrates and the presence of PCDD/Fs and PCBs. The concentrations of chemical contaminants in animal products consumed by humans such as eggs were estimated using transfer models when animals were fed with BSFL reared on unauthorized substrates. The concentrations were below the maximum limits in all cases
Territoria Naturalia. Aménager sans altérer, entre partage et réserve
No full textInternational audienc
La sobriété numérique comme levier d’innovation pour les entreprises : une analyse sur données d’enquête
No full textInternational audienceThis article examines whether digital sobriety practices implemented by companies could act as a barrier to innovation. As investment in digital technology and equipment is a key factor in driving innovation, it is important to understand the implications of adopting more restrained digital practices on a company's innovation capabilities. To investigate this, we draw on a survey of 2,875 companies in Brittany and Pays de la Loire. Our findings reveal that companies most committed to digital sobriety are also those that innovate the most in terms of processes and products, although the positive relationship is more significant for process innovations. Thus, implementing responsible practices, particularly in digital areas, is not an obstacle to innovation and may even help companies move toward more sustainable organizational models.Cet article s’intéresse à la question de savoir si les actions de sobriété numérique mises en œuvre par les entreprises peuvent constituer ou non un frein à l’innovation. L’investissement dans les technologies et équipements numériques étant un facteur clé de l’innovation, il est important de comprendre les implications d’une plus grande sobriété numérique sur les capacités d’innovation des entreprises. Pour cela, nous mobilisons une enquête réalisée auprès de 2875 entreprises en Bretagne et Pays de la Loire. Nos résultats montrent que les entreprises les plus engagées dans des démarches de sobriété numérique sont aussi celles qui innovent le plus en procédé et en produit, même si la relation positive est plus significative pour les innovations de procédé. La mise en œuvre de pratiques responsables, notamment en matière de numérique, n’est donc pas un obstacle à l’innovation et peut même permettre aux entreprises d’aller vers des organisations plus durables. CODES JEL : L2, O3
Airbnb and Returns to Housing Capital: Evidence on Inequality from Administrative Data
No full textInternational audienc
Efficient Memory Usage For Edge FaaS Platforms
No full textInternational audienceFunction as a Service (FaaS) is a great fit for data and event processing in Edge environments. These environments are characterized by resource-constrained devices that require efficient memory usage optimizations. Existing optimization so- lutions for memory in high-end clusters such as data centers cannot be used in Edge environments because they either depend on unavailable hardware features such as RDMA or require resource-intensive analysis such as periodic memory scanning, compression, or decompression. In this paper, we introduce Extensible RUNtimes (ERUN), a lightweight mechanism for existing FaaS runtimes aimed at optimizing memory utilization by reducing the memory usage of idle sandboxes (i.e., those awaiting function execution). ERUN operates through two main actions: Shrink and Expand. The Shrink operation unloads libraries and reclaims memory from the sandboxes, while the Expand operation quickly reloads the libraries when a function is executed. The Expand operation leverages an in-memory store that maintains a single instance of discarded libraries on the node. We implement the ERUN mechanism, which can be applied to any runtime environment, in a Python runtime. We extensively evaluated our prototype in a 10-node Edge cluster using 10 popular FaaS functions. The results show that ERUN can reduce idle sandbox memory usage by up to 23.13× and improve the warm-start ratio by 1.38×, while incurring less than 2% overhead on function execution time and energy usage
Analysis of infectious complications in paediatric autoimmune neutropenia: a French nationwide retrospective cohort study
No full textInternational audienceBackground: Autoimmune neutropenia (AIN) is the main cause of chronic neutropenia in children, but its infectious consequences remain poorly studied. The primary objective of this study was to evaluate infectious events leading to emergency department or hospital admissions during the first 2 years following the diagnosis of AIN in children.Methods: We performed a retrospective, multicentre analysis of medical records from 21 French university hospitals of patients aged under 18 years diagnosed with AIN with positive antineutrophils autoantibodies. We collected data on emergency room visits and hospitalisations in the 2 years following diagnosis, causes of these events, microbiology results, management and outcome.Results: One hundred and sixty-eight patients were enrolled. Median age at diagnosis of AIN was 13 months. AIN was predominantly diagnosed during an infectious episode (n=120, 71%). In the 2 years of follow-up after diagnosis, 248 events of emergency room visits and/or hospitalisations were reported (0.77 per patient-year). The most frequent diagnoses were common childhood viral or bacterial infections. The incidence rate of severe infections was 0.003 per patient-year. Despite the predominance of viral infections, 177 episodes (71%) led to hospitalisation and 166 (68%) to the initiation of antibiotic therapy, for a median duration of 7 days (IQR 3-10).Conclusion: The risk of severe infections in children with AIN is low. During follow-up, we suggest being attentive to signs of severity during fever, particularly in children over 3 years of age and/or with other immunological comorbidities but not proposing systematic hospitalisation or additional antibiotic therapy