Three device families and two decades of SPAD and SiPM filings
Single photon detector technology in this dataset spans three primary device families: SPAD-based pixel arrays, SiPM (Silicon Photomultiplier) arrays composed of Geiger-mode avalanche photodiodes, and scintillator-coupled detector modules for ionizing radiation sensing. All three exploit avalanche multiplication — reverse-biasing a p-n junction above its breakdown voltage — to generate a measurable electrical signal from a single photon impact. The dataset covers 18 distinct assignees across CN, US, JP, KR, DE, IT, ES, FR, WO, IL, and BR jurisdictions, with publication dates ranging from 2003 to October 2025.
Key technical dimensions tracked across the dataset include quenching circuits (passive, active, and gated architectures to terminate and reset avalanche events), readout electronics (time-to-digital converters, digital SiPM architectures, and event-driven pixel circuits), and system integration approaches — from monolithic CMOS integration of SPADs with readout logic to metalens focusing structures and neuromorphic triggering circuits. Performance metrics of interest include detection efficiency (PDE), dark count rate (DCR), timing jitter, afterpulsing probability, and dynamic range. According to WIPO, solid-state photon sensing is among the fastest-growing technology categories in global patent filings.
In Geiger mode, a photodiode is reverse-biased above its breakdown voltage. A single photon impact triggers a self-sustaining avalanche of carriers, producing a large, standardised digital pulse. A quenching circuit — passive, active, or gated — terminates the avalanche and resets the device for the next detection event. This binary, photon-counting behaviour distinguishes SPADs from linear-mode APDs, which produce an analog signal proportional to photon flux.
Single photon detectors in the 2026 technology landscape span three primary device families — SPAD-based pixel arrays, SiPM arrays, and scintillator-coupled detector modules — all exploiting avalanche multiplication above breakdown voltage to detect individual photons. The dataset covers 18 distinct assignees across 11 jurisdictions, with publication dates from 2003 to October 2025.
From gated SPADs to neuromorphic arrays: the innovation timeline
Single photon detector patent activity divides into four distinct phases, with a clear acceleration in filings from 2018 onward. The earliest records in this dataset — dating to 2003 — concern gated SPAD operation and precision photon arrival time measurement. Hioki E.E. Corporation’s 2003 Japanese filing established gated biasing above breakdown for synchronised single photon detection. Politecnico di Milano’s 2006 German filing introduced output circuit differentiation for sub-nanosecond timing precision. Koninklijke Philips Electronics’ 2008 Chinese filing established the digital SiPM architecture with per-cell digital logic for TOF-PET applications — a design that would anchor the medical imaging cluster for the following decade.
The 2010–2018 consolidation period is characterised by medical imaging refinements and SiPM energy resolution optimisation. Koninklijke Philips filed multiple patents in this window — including on silicon photomultiplier energy resolution, improved light detection in pixelated PET detectors, and virtual PET detector and quasi-pixelised readout schemes — reflecting intensive investment in scintillator-SiPM coupling geometry and TOF-PET timing. Siemens’ 2012 Chinese filing proposed sub-millimeter APD arrays as an alternative to SiPM nonlinearity in PET applications.
From 2018 onward, automotive and consumer sensing applications began driving filings. Waymo’s 2020 Israeli filing and its 2022 Korean equivalent demonstrate monolithic SiPM and linear-mode APD co-integration for LiDAR. Chinese assignees — including Shenzhen Guangwei Technology, Shandong University, and Southeast University — began appearing in volume from 2019. The most recent filings (2023–2025) reveal three emerging directions: neuromorphic SPAD arrays, event-driven imaging, and AI-assisted photon acquisition. According to IEEE, event-driven sensing architectures are among the most actively published topics in solid-state circuits research.
“The most recent filings reveal a structural shift: the next generation of single-photon imagers will not be framed around pixel arrays but around asynchronous event streams.”
Four technology clusters driving the single photon detector patent landscape
The patent landscape organises into four coherent technology clusters, each with distinct device physics, application targets, and assignee profiles. SPAD-based Geiger-mode arrays with digital readout represent the dominant architecture across the dataset; SiPM architectures for radiation detection dominate the medical and nuclear security sub-fields; gated and active quenching circuits for near-infrared SPADs address quantum communications; and a fourth emerging cluster brings neuromorphic, event-driven, and hybrid detector integration to the fore.
Cluster 1: SPAD-Based Geiger-Mode Arrays with Digital Readout
SPADs operate above breakdown voltage in Geiger mode, generating a binary digital pulse per photon detection event. Arrays are organised into macropixels with shared or per-cell time-to-digital converters (TDCs) for time-of-flight measurement. XO Semiconductor’s 2025 Japanese filing covers pulse counting from SPAD arrays with global clock synchronisation for image quality determination. Shenzhen Guangwei Technology’s 2019 Chinese filing introduced coincidence detection requiring at least 3 of 8 SPADs to fire in a macropixel, suppressing ambient light false triggers in TOF ranging. Southeast University’s 2024 Chinese filing describes a differential TOF measurement approach that preserves lateral resolution while filtering noise in SPAD array readout.
Cluster 2: Silicon Photomultiplier Architectures for Radiation Detection
SiPMs aggregate many Geiger-mode microcells in parallel on a single substrate to produce an analog sum proportional to photon flux. This cluster dominates medical imaging (PET, SPECT, CT) and nuclear security applications. Koninklijke Philips’ 2012 Chinese filing describes per-cell digital circuits with trigger logic for sub-nanosecond timing in TOF-PET. FLIR Detection (Teledyne FLIR)’s 2018 Chinese filing covers scintillator-coupled SiPM arrays with energy windowing ADC and temperature-stabilised bias for portal radiation monitoring. Koninklijke Philips’ 2024 Chinese filing targets sub-100 ps coincidence time resolution in PET by detecting prompt Cherenkov photons from a semiconductor crystal combined with direct-conversion charge readout.
Explore the full SPAD and SiPM patent dataset — search by assignee, jurisdiction, and technology cluster.
Search patents in PatSnap Eureka →Cluster 3: Gated and Active Quenching Circuits for Near-Infrared SPADs
InGaAs/InP SPADs operating in the near-infrared (1310 nm and 1550 nm telecom bands) require specialised quenching to manage afterpulsing. Shandong University filed three patents (2020–2022) specifically on InGaAs SPAD quenching circuits for quantum key distribution, representing a concentrated national academic effort. The 2021 filing describes an FPGA-controlled hybrid gated-passive plus active quench architecture suppressing higher-order afterpulse cascades. China Electronics Technology Group Corporation’s 2021 filing covers RF transformer-based large-amplitude gate pulses with short rise and fall times for silicon SPAD in gated mode.
Cluster 4: Neuromorphic, Event-Driven, and Hybrid Detector Integration
Emerging filings from 2022–2025 introduce computation-in-pixel and AI-native paradigms, treating SPAD arrays as asynchronous event generators rather than frame-based imagers. The Commonwealth of Australia’s 2025 European Patent Office filing implements neuron-emulating trigger circuits across overlapping receptive fields, forming a convolution layer directly on the SPAD array. Portland State University’s 2025 WO filing formalises per-pixel inhibition based on photon flux estimation to suppress readout power in low-activity pixels — directly targeting always-on mobile and wearable sensor deployment. IRIDAE S.R.L.’s 2023 Italian filing replaces conventional ASIC signal processing with neural network-based gamma photon event reconstruction.
Shandong University filed 3 patents between 2020 and 2022 focused on InGaAs/InP near-infrared SPAD quenching circuits for quantum key distribution at 1310 nm and 1550 nm telecom bands, representing a concentrated Chinese national academic effort in QKD-adjacent single photon detector subsystems.
Application domains: LiDAR, PET imaging, quantum key distribution, and life sciences
Single photon detector applications in this dataset span six distinct domains, each with characteristic device requirements and leading assignees. Automotive LiDAR and 3D sensing represents the highest-activity commercial sector in recent filings, while medical imaging (PET, SPECT, CT) constitutes the most historically dense cluster.
Automotive LiDAR and 3D Sensing
SPAD and SiPM arrays are used for time-of-flight ranging in autonomous vehicles and consumer depth cameras. Waymo’s 2022 Korean filing demonstrates monolithic SiPM and linear-mode APD (LmAPD) co-integration for high-dynamic-range LiDAR — addressing the fundamental tension between single-photon sensitivity at long range and saturation avoidance at short range. Semiconductor Components Industries (onsemi) filed a 2021 Chinese patent on a periodic histogramming TDC architecture minimising per-pixel circuit area in global-shutter SPAD LiDAR, and a 2025 Chinese filing on per-SPAD digital enable and disable for performance-degraded microcells. OmniVision Technologies’ 2023 Chinese filing describes a digital-analog hybrid output for bright-ambient LiDAR detection.
Medical Imaging: PET, SPECT, and CT
The most historically dense cluster in this dataset is dominated by Koninklijke Philips across both analog and digital SiPM architectures. Philips’ 2008 foundational digital SiPM detector pixel design for TOF-PET has been followed by filings on energy resolution, pixelated detector light detection, and virtual detector geometries. The University of Texas System’s 2018 Chinese filing describes a cost-reduced depth-of-interaction PET approach using fewer SiPMs via dichotomous readout. JOB Corporation’s 2025 German filing covers a columnar scintillator assembly with optical connector for CT photon counting. The critical performance barrier for whole-body TOF-PET — sub-100 ps coincidence time resolution — is targeted by Philips’ 2024 Cherenkov photon detection filing.
High-Energy Physics and Scientific Instrumentation
Fermi National Accelerator Laboratory’s 2023 publication describes an ARAPUCA-based SiPM photon detection system for a liquid argon neutrino detector with near-4π coverage, developed for the DUNE low-energy physics programme. Jagiellonian University’s 2015 Spanish filing covers a matrix detector for spatial-temporal reconstruction of annihilation events in PET. These filings reflect the deep institutional overlap between medical physics instrumentation and particle physics detector development, as documented by Nature in reviews of scintillator-coupled photodetector advances.
Quantum Communications and Spectroscopy
Shandong University’s 2022 Chinese filing covers InGaAs SPAD operation at 1310 nm and 1550 nm telecom bands for quantum key distribution. The State University of New York Research Foundation’s 2010 Chinese filing describes a 32-channel PMT-based fiber-coupled single-photon spectrometer with greater than 20-bit dynamic range, greater than 10⁸ photon counts per second, and 3,300 frames per second for DNA sequencing applications.
Flow Cytometry, Life Sciences, and Nuclear Security
Life Technologies Corporation (Thermo Fisher Scientific)’s 2025 Japanese filing covers SPAD array-based fluorescence detection in flow cytometry with activated-pixel count correction. FLIR Detection’s 2016 Chinese filing and Arktis Radiation Detectors’ 2025 Chinese filing address SiPM scintillator portal monitoring and compact panel detection of ionizing radiation respectively — reflecting the displacement of PMT-based radiation monitors by solid-state alternatives.
Medical imaging, nuclear monitoring, spectroscopy, and scientific instrumentation all show active transition from photomultiplier tubes to solid-state alternatives. R&D teams entering these markets should treat SiPM or SPAD as the baseline detector technology rather than PMT, as PMT-based approaches are now the exception rather than the rule in recent patent filings.
Geographic and assignee concentration: China leads filings, Philips leads volume
China accounts for approximately 40% of total records in this dataset — reflecting both domestic R&D activity and China as a filing jurisdiction for multinational assignees. Japan represents roughly 15% of records, while Germany, South Korea, the United States, and international (WO/EP) filings each contribute 5–10%. Italy, France, Spain, Israel, Brazil, and Mexico each contribute isolated filings. Innovation is notably concentrated in a small number of assignees in medical imaging, with broader fragmentation in LiDAR and consumer imaging.
Koninklijke Philips N.V. is the most prolific assignee in this dataset with at least 9 distinct records spanning 2008–2024. OmniVision Technologies holds at least 5 filings from 2023–2025 in CN jurisdiction covering event-driven pixel architectures and dual-mode SiPM for LiDAR. Shandong University has 3 filings (2020–2022) focused on InGaAs/InP near-infrared SPAD quenching circuits for QKD. Ji Hua Laboratory (Jihua Laboratory) has 3 filings (2024–2025) on measurement methods for single-photon detector efficiency, afterpulse probability, dead time, and timing jitter — indicating Chinese national metrology investment. Waymo holds 2 filings (2020–2022) on monolithic SiPM and LmAPD combination arrays for LiDAR dynamic range extension.
Chinese institutional activity is distributed across universities and research institutes — Shandong University, Southeast University, Tianjin University, Ji Hua Laboratory — rather than consolidated in single commercial entities. This contrasts with the medical imaging sub-field, where Philips’ dominance is pronounced. IP strategists should monitor CN filings actively and assess freedom-to-operate in QKD-adjacent SPAD subsystems, as noted in OECD analysis of China’s expanding role in deep-tech patent portfolios.
Koninklijke Philips N.V. is the most prolific assignee in the single photon detector patent landscape with at least 9 distinct records spanning 2008–2024, covering digital SiPM architecture, TOF-PET optimisation, energy resolution, virtual detector geometries, and quality control methods for pixelated PET detectors — with filings primarily into CN and JP jurisdictions.
Map assignee IP portfolios and identify white-space opportunities across SPAD and SiPM technology clusters.
Analyse assignees in PatSnap Eureka →Five frontier directions reshaping single photon detector technology in 2025–2026
Five frontier directions are discernible from filings dated 2023–2025 in this dataset. Taken together, they represent a structural shift away from frame-based, analog-sum detector architectures toward intelligent, asynchronous, and energy-aware sensing paradigms.
1. Neuromorphic SPAD Architectures
The Commonwealth of Australia’s 2025 European Patent Office filing implements overlapping receptive fields with silicon neurons, enabling on-chip convolution and event-triggered readout. This eliminates frame-based acquisition entirely, reducing data bandwidth by orders of magnitude for machine vision applications. The design treats the SPAD array not as a camera but as a spiking neural network front-end.
2. Energy-Efficient Pixel Inhibition and Selective Readout
Portland State University’s 2025 WO filing formalises per-pixel inhibition based on photon flux estimation, suppressing power consumption in dark-scene pixels. This directly targets always-on mobile and wearable sensor deployment — a use case where the power budget of conventional SPAD readout architectures has been prohibitive.
3. Event-Based Processing of SPAD Photon Streams
Ubisept, Inc.’s 2025 Korean filing describes exponential moving average-based flux estimation with event triggering, enabling SPAD arrays to mimic dynamic vision sensors (DVS). This convergence of single-photon sensitivity with neuromorphic event representation is a significant architectural shift — SPAD arrays gain the temporal resolution of event cameras while retaining photon-counting sensitivity.
4. Hybrid Direct-Conversion and SPAD Detector for Cherenkov-Based TOF-PET
Koninklijke Philips N.V.’s 2024 Chinese filing describes a SPAD array detecting prompt Cherenkov photons from a semiconductor crystal, combined with direct-conversion charge readout — targeting sub-100 ps coincidence time resolution in PET. Sub-100 ps timing is a critical barrier for whole-body TOF-PET and represents one of the most demanding performance specifications in the dataset.
5. Quantum Dot and Flat-Lens Integration for NIR-Extended SPAD
Infineon Technologies AG’s 2023 Chinese filing and Margin Technology (Zhuhai)’s 2023 Chinese filing both demonstrate material-level and optical-element-level enhancements to extend sensitivity and fill factor beyond silicon SPAD limits. Infineon’s quantum dot photodetector approach and Margin Technology’s integrated metalens structure for 0.85 µm avalanche diodes represent two complementary strategies for pushing SPAD spectral response into the near-infrared without switching to InGaAs substrates.
“LiDAR dynamic range remains an unsolved integration challenge — Waymo’s monolithic SiPM and LmAPD approach and onsemi’s per-microcell digital enable architecture both address the tension between single-photon sensitivity at long range and saturation avoidance at short range, with no dominant architectural solution yet established.”
Ji Hua Laboratory (Jihua Laboratory) in China filed 3 patents in 2024–2025 covering measurement methods for single-photon detector efficiency, afterpulse probability, dead time, and timing jitter — signalling the emergence of performance characterisation and calibration as a distinct IP sub-domain within the single photon detector landscape.