A new review by Jean-Marc Bertho of the French Nuclear Safety and Radiation Protection Authority, published in Frontiers in Nuclear Engineering, takes stock of this movement three decades on. The review traces how citizen-led measurement of radioactivity evolved from a post-accidental coping mechanism into a structured citizen science practice, and it identifies the technical, social and institutional challenges that currently limit the sustainability and useability of citizen measurement data. It also charts the opportunities, from artificial intelligence-based data cleaning to official accreditation, that could finally integrate citizen detectors into environmental monitoring and emergency decision-making. The stakes are considerable: the Safecast network alone has collected more than 250 million measurements worldwide, and the OpenRadiation database in Europe holds more than 1.6 million results, with roughly 100,000 added per year.

The basic architecture of every citizen measurement network is strikingly simple. Citizens are provided with detectors, helped to use them, and given means for publishing and sharing their results, which translates into a system with three components: a detector, a communication tool eventually monitored by the citizen, and an interactive map for publishing measurement results. What separates the networks is detector complexity. Autonomous detectors incorporate GPS chips, internal clocks, GSM communication modules and temperature sensors, and some models include solar panels so they can operate as fixed, energy-independent beacons; this is the approach of Safecast detectors adapted for the SaveEcoBot network in Ukraine, mainly for users’ safety reasons. The second family is deliberately minimal: a detector tube, associated electronics, a battery and a Bluetooth module controlled by a smartphone application. Because metadata such as GPS position, time, photographs and literal descriptions of the environment are collected from the smartphone, these detectors are much cheaper and can capture a richer array of contextual information with each measurement.

All of these networks operate according to the European Citizen Science Association’s ten principles, and one fundamental principle underpins the entire enterprise: openness. Data are freely accessible, but so are the source codes of applications and websites and the electronic diagrams of detectors, so that anyone can reproduce the whole system or extract the whole or part of the database for specific analyses. The sole limitation is citation of sources under CC-BY licence requirements. This openness is essential to maintain the trust of citizens engaging in the networks. Beside dedicated detectors, smartphone applications exist that use the CMOS camera chip to detect ionizing radiation by lighting up pixels with excess charge when the camera is obturated into full dark mode. However, the sensitivity and specificity of these CMOS cameras to ionizing radiation is low, especially at dose rates below 0.5 microsieverts per hour, and a specific calibration curve is needed for each camera type. Such applications are not recommended in areas with a usual background, although with noise-removing functions and calibration they may help detect cosmic rays during commercial flights.

Reliability is the central technical question. Most existing networks use Geiger-Mueller technology, which is robust and reliable provided the tubes are tested before use. The OpenRadiation network, for instance, selects only GM tubes falling within a tolerance range of approximately 10 percent of a reference dose rate; a recent campaign showed that the 366 selected tubes provided a response within plus or minus 6.2 percent of the reference value, with a rejection rate of 86 percent. Other networks, such as FIfF e.V./TDRM in Germany, use CsI(Tl) or NaI scintillator crystals to register spectrometric profiles capable of identifying radioactive releases from nuclear power plants. In all cases, the European networks are associated with expert institutes that test and characterize detector response. Very few intercomparison studies with citizen detectors have been published, but a 2019 field campaign in an area with locally high concentrations of naturally occurring radionuclides, mapped first with a high-precision NaI detector and referenced with an ionisation chamber, demonstrated that citizen GM detector results were comparable to professional gas detectors at five out of six field points. At the point with the highest dose rate, a large variability was observed, hypothesized to result from detector heating above 50 degrees Celsius in the rising sun, which induced aberrant responses.

A longer-term comparison conducted in Japan between 2019 and 2023 pitted the Kyoto University radiation mapping system KURAMA against the Safecast citizen mapping system. The study showed that KURAMA’s geographical coverage is larger, but that Safecast was able to catch local situations, such as residential areas and public facilities, that KURAMA does not monitor. Uncorrected Safecast data exhibited a systematic bias, although regression analysis revealed a strong linear correlation between the two datasets, supporting the concept that citizen measurement data might be useful if integrated into environmental monitoring and policymaking. From the metrological point of view, citizen results are already reliable: measurement uncertainty is managed so that it stays below 15 percent, which requires a minimum of 50 detected counts per measurement. Safecast’s large pancake-format gas detector, integrated over one minute, detects more than 50 counts except in very low background areas, while OpenRadiation prevents a measurement from being recorded until at least 50 counts have accumulated and recommends end-users to accumulate as many counts as possible to improve accuracy.

Yet reliability of the instrument is only half the problem; reliability of the measurement report is the other half. Recommended protocols, such as measuring at one metre above the ground, are generally well followed, and some contributors spontaneously developed wooden detector holders to control measurement height, demonstrating that citizens can develop skills and practical tools to ensure protocols are respected. But the fundamental principle of citizen science is that people are free to measure what they want, where they want, when they want and as they want, governed by their own primary interests. A 2018 study in Belarus gave OpenRadiation detectors to 17 pupils for a month, yielding 645 measurements mostly in the range of 0.05 to 0.18 microsieverts per hour, with three hot spots identified, each associated with fireplaces burning locally contaminated wood. The GPS timeline revealed that most pupils made their first measurement at home, underscoring the importance of the home as a safe place, and that each participant organized measurement places differently, from tight grids covering the nearest neighbourhood to walks around the settlement. This individual variability, along with weather, background variations and naturally radioactive objects such as granite street furniture, constitutes a real source of uncertainty that networks manage through moderation, either before publication, as the Citistra network does with Safecast detectors, or afterwards, as OpenRadiation does by labelling any measurement above 0.3 microsieverts per hour as atypical and adding explanatory comments rather than removing results.

The deeper difficulty for networks other than SaveEcoBot is the absence of a radiological issue to sustain interest and funding. These networks orient themselves towards educational activities and risk awareness-raising, which are not sufficient to maintain citizen engagement or obtain long-term financial support, especially since their intended emergency function, providing real-time field data in a radiological crisis, is fortunately rarely needed. Contribution patterns follow the classic citizen science curve: a very limited number of users provides the majority of measurements, while the vast majority publish occasionally or only once, encouraged by dose rate values that vary in a narrow band between 0.05 and 0.15 microsieverts per hour in most cases. Motivation studies classify participation drivers as intrinsic, linked to personal interest in environment, health and science, and extrinsic, linked to recognition and reputation, and underline the importance of belonging to a community. A second kind of activity, the use of the database by external users, could represent a powerful motivation when communicated back to contributors, but first requires delineating the useability of the databases in view of the identified uncertainties.

A detailed quality study of the OpenRadiation database between April and December 2024 shows just how tractable those uncertainties are. Of 75,505 published measurements, 30,613 generated alerts above the 0.3 microsievert per hour threshold, but 96.01 percent of these alerts were made on board a plane, at altitudes systematically over 3,000 metres, explaining dose rates up to 6.5 microsieverts per hour as a simple mistake in the measurement mode setting. Visual map inspection attributed 128 measurements to areas with high natural background and 63 to areas near the Fukushima Daiichi plant, illustrating the behaviour of contributors seeking hot spots. Contact with contributors explained the remaining alerts, including 22 measurements near urban objects enriched in naturally occurring materials and 979 measurements from just four contributors testing reference radioactive sources. The proportion of published measurements above the threshold that remains unexplained is a mere 0.022 percent. The review also documents a striking demonstration of complementarity: in 2023 a contributor measured high radioactivity on a beach in western France, with ambient dose rates between 0.4 and 2.63 microsieverts per hour and a hot spot in the cliff at 78.4 microsieverts per hour, while the nearest official Teleray beacons sit more than 30 kilometres away. An expert study by ASNR traced the anomaly to monazite, a red crystal enriched in thorium and uranium producing red sands through erosion, and showed an absence of risk for the population.

The landscape is now changing rapidly, driven in part by the war in Ukraine, which re-activated fear of nuclear weapons use in the battlefield, and by the sheer volume of data these networks have accumulated. SaveEcoBot, created to provide environmental radiation monitoring during the Russian war in Ukraine, has become the first citizen network in the world officially accredited for radiation environmental monitoring. Research projects such as CosmicOnAir, which tracks solar flares through citizen measurements during flights, and CORALE, which uses travellers to fill gaps in the radiological exposome of the French CONSTANCES epidemiological cohort, demonstrate that citizen data are fully useable for science. Two new European projects aim to institutionalize this integration: CIThARA, started in January 2026, will develop artificial intelligence-based tools to clean and assign confidence indices to the 1.3 million results in the OpenRadiation database, of which more than 90 percent proved useable after cleaning, and to integrate them with official networks such as Teleray in France and RESORAD in Italy, while ORION, starting in July 2026, will issue recommendations on how to use citizen and drone measurements in emergency decision-making. Crucially, the review insists that citizen networks must remain independent from public authorities, since that independence is an essential driver of citizen confidence, and that the use of their data should be seen as nothing more than a recognition of its quality. Dialogue between citizens, technical support organizations and regulators, the review concludes, is the key to future developments, with small detectors and even do-it-yourself kits remaining at the core of an educational mission that builds the population’s resilience should an accident ever occur.

Subject of Research: Citizen science radiation measurement networks for ambient dose rate monitoring

Article Title: Citizen measurement of ambient dose rate: challenges and opportunities

Article References: Bertho, J.-M. (2026). Citizen measurement of ambient dose rate: challenges and opportunities. Frontiers in Nuclear Engineering, 5, Article 1878296. https://doi.org/10.3389/fnuen.2026.1878296

Image Credits: AI Generated

DOI: 10.3389/fnuen.2026.1878296

Keywords: citizen science, ambient dose rate, radiation measurement, Geiger-Mueller detector, Safecast, OpenRadiation, environmental monitoring, nuclear emergency preparedness, radiation protection, Fukushima, Chornobyl, data uncertainty

Brenda Betancourt, a Term Associate Professor of Statistics in George Mason’s College of Engineering and Computing, has received $286,130 from the U.S. Food and Drug Administration for a project titled Regulatory Grade Frameworks for Assessing Record Linkage Uncertainty in Real-World Evidence. The funding began in September 2026 and runs through September 2028. Over those two years, Betancourt will evaluate how uncertainty in record linkage affects the reliability, interpretability, and evidentiary confidence of analyses based on linked real-world data that are used to support FDA regulatory decision-making. In other words, the project asks a deceptively simple question: if the links between records are not certain, how certain can the conclusions drawn from them be?

The stakes are considerable. Record linkage is the invisible plumbing of real-world evidence. When a researcher wants to know whether patients exposed to a medication later experienced a particular adverse event, the exposure records typically live in one database, perhaps pharmacy claims, while the outcome records live in another, perhaps a hospital discharge registry. Joining those datasets requires deciding which rows refer to the same individual, often without a shared unique identifier such as a national patient number. Probabilistic matching algorithms compare names, dates of birth, addresses, and other fields, assigning scores that indicate how likely two records are to describe the same person. Every one of those assignments carries a probability of error, and those errors propagate silently into every downstream estimate.

Statisticians have long understood that linkage errors can bias results in either direction. False matches, in which records from two different people are merged, can dilute genuine signals or manufacture spurious associations. Missed matches, in which records from the same person remain split, can fragment a patient’s medical history and undercount exposures or outcomes. The direction and magnitude of the bias depend on the nature of the analysis, the matching variables, and the population being studied. What has been missing, particularly in the regulatory context, is a standardized framework that quantifies this uncertainty in a way that meets the evidentiary bar regulators demand when a decision affects public health.

That is the gap Betancourt’s project is designed to fill. The phrase regulatory grade in the project title signals the ambition: not merely to measure linkage uncertainty in an academic sense, but to develop frameworks rigorous and transparent enough to be used in submissions that inform FDA decisions. Regulatory science operates under constraints that academic research often does not. Methods must be reproducible, their assumptions explicit, their outputs interpretable by reviewers who must weigh evidence under statutory standards. A framework that cannot communicate how much of a reported effect might be attributable to linkage error is of limited use to an agency deciding whether to approve, restrict, or warn about a therapy.

The mathematical core of the problem is rich. Probabilistic linkage models, descendants of the classical Fellegi-Sunter framework, estimate the likelihood that a pair of records matches based on agreement and disagreement patterns across comparison fields. More modern approaches bring in Bayesian modeling, clustering, and machine learning classifiers, each producing not a single deterministic assignment but a posterior distribution over possible linkages. The challenge is propagation: how does uncertainty in those assignments flow through data cleaning, cohort construction, statistical modeling, and ultimately into confidence intervals and p-values for the regulatory question at hand? Ignoring that propagation can yield analyses that appear far more precise than they actually are, a phenomenon statisticians describe as understated uncertainty.

Several strategies exist in principle for handling this propagation. One is to condition on the most probable linkage and then adjust estimates using sensitivity analyses that vary the assumed error rates. Another is multiple imputation over plausible linkages, generating several alternative versions of the linked dataset and combining the resulting estimates so that the final inference reflects linkage ambiguity. A third embeds the linkage model and the analysis model in a joint Bayesian framework, allowing uncertainty to be integrated out rather than fixed. Each approach carries computational costs and modeling assumptions, and each performs differently depending on data quality, record volume, and the discriminative power of the matching fields. A central task for the project is evaluating when these methods deliver trustworthy inference and where they fall short.

The timing of the work reflects a broader shift in how evidence reaches regulators. The FDA has invested heavily in real-world evidence programs, including the RWE framework established under the twenty-first Century Cures Act, which directed the agency to clarify how real-world data can support approvals and label changes, particularly in areas such as rare diseases, oncology, and post-market safety surveillance where randomized trials are impractical or unethical. As the volume of linked real-world analyses grows, so does the need for standards governing every step of the pipeline. Linkage is among the earliest and least visible of those steps, which makes its errors especially insidious: they are baked in before most analysts ever see the data.

For patients and clinicians, the practical payoff of this research is confidence. If a safety signal emerges from linked claims and hospital data, clinicians and patients deserve to know whether that signal could plausibly be an artifact of mismatched records. Conversely, if a real-world analysis finds no elevated risk, regulators should know how robust that null finding is to plausible linkage errors. Quantified linkage uncertainty turns these questions from matters of intuition into matters of measurement, allowing reviewers to weigh evidence with a clearer sense of its fragility or strength. In an era when real-world evidence increasingly shapes drug labels, coverage decisions, and clinical guidelines, that clarity is not a technical luxury but a foundation of trustworthy medicine.

The project also highlights the growing role of statisticians in regulatory science and the position of universities near the federal policy apparatus in shaping it. George Mason, Virginia’s largest public research university, enrolls more than 40,000 students and sits near Washington, D.C., placing its researchers in close proximity to the agencies whose standards define the evidentiary landscape. Over the next two years, Betancourt’s work will contribute to a question that touches every linked dataset behind every regulatory submission: how sure can we be that the records we joined belong together, and how should that residual doubt be reflected in the evidence we act upon. The answer may help determine how much trust real-world evidence ultimately earns.

Subject of Research: Quantifying record linkage uncertainty in real-world evidence for FDA regulatory decision-making

Article Title: Betancourt studying regulatory grade frameworks for assessing record linkage uncertainty in real-world evidence

Article References: Betancourt studying regulatory grade frameworks for assessing record linkage uncertainty in real-world evidence. (n.d.). Original publication

Image Credits: AI Generated

DOI: Not provided

Keywords: record linkage, real-world evidence, FDA, regulatory science, statistics, data uncertainty, George Mason University, probabilistic matching, pharmacovigilance, electronic health records, Bayesian methods, drug safety