16 Years of Patent Filing: What the Volume Data Reveals
Bosch filed 11,745 automotive sensor and ADAS patents between 2010 and 2026 — a volume that places the company among the most active innovators in the global automotive supplier landscape. The filing trajectory is not linear: activity accelerated sharply from 567 patents in 2016 to 922 patents in 2018, reflecting the mass-market inflection point for Level 2 ADAS deployment. Peak output arrived in 2023 with 1,172 patents, followed by 1,074 patents in 2024.
The 40% pending rate — representing 4,679 applications — is a particularly telling signal. It indicates aggressive recent filing activity and suggests that a large tranche of Bosch’s most current technical work has not yet been granted or published in full. Of the remaining portfolio, 3,440 patents (29%) are active and 2,622 (22%) are inactive or abandoned. The high active-to-abandoned ratio reflects strong prosecution discipline in mature markets.
Bosch filed 11,745 automotive sensor and ADAS patents between 2010 and 2026, reaching peak annual output of 1,172 patents in 2023 and 1,074 patents in 2024, with 40% of the total portfolio currently in pending status.
It is worth noting that the 2025 and 2026 patent counts will be materially underreported in current databases due to the standard 18-month publication lag between filing and public disclosure. Web intelligence confirming active product launches through 2026 — including a new 8MP multi-purpose camera for China and a next-generation radar entering serial production — indicates that actual filing activity remains strong.
Three-Phase Technology Evolution: From Sensors to Software
Bosch’s ADAS patent portfolio maps cleanly onto three distinct strategic phases, each defined by a shift in where the company placed its technical bets. Understanding these phases is essential for interpreting both the current patent landscape and the 2026 product roadmap.
Phase I — Foundation Building (2010–2016)
The first phase centred on sensor reliability and basic ADAS functions. Key innovations included functional testing methods for ultrasonic sensors using secondary-mode frequencies — enabling concurrent testing without disrupting live driver assistance operations — and safe protocols for activating and deactivating autonomous mode via steering wheel button combinations with timing constraints. Patent activity grew steadily, reaching 567 patents in 2016 before accelerating sharply.
Phase II — Multi-Sensor Integration (2017–2021)
The second phase focused on sensor fusion and Level 2 automation. Bosch developed autonomous parking systems that operate without external infrastructure dependency, integrated ultrasonic sensors with radar and camera systems for close-range velocity assessment in adaptive cruise control, and pioneered methods for dynamically adjusting sensor usage and measurement rates based on real-time trajectory quality monitoring. This last innovation — documented in patent US12128890B2 — demonstrated energy efficiency gains through adaptive sensor operation. Patent output surged from 922 in 2018 to 1,097 in 2021.
Safety of the Intended Functionality (SOTIF), standardised under ISO/PAS 21448, addresses hazards arising from performance limitations or foreseeable misuse of a system — distinct from the hardware or software failures covered by ISO 26262 functional safety. Bosch’s quantitative SOTIF method (US12330693B2) enables formula-based failure rate calculation using driver status events and system fault severity, bridging both standards.
Phase III — AI-Driven Autonomy (2022–2026)
The third phase marks Bosch’s pivot to software-defined vehicles, Level 3+ capabilities, and edge-cloud architecture. Four innovation clusters define this period. First, Bosch became the first Tier 1 supplier to develop a complete radar system including a proprietary in-house system-on-chip (SoC), using RF CMOS technology with 22nm transistors — a single chip integrating high-frequency RF and digital circuits for compact footprint with high computing power. Second, AI and machine learning were embedded throughout the perception stack, from a visual analytics platform for improving object detection models (US11587330B2, valued at $330,000) to multi-task CNN architectures for parallel perception tasks. Third, a quantitative SOTIF analysis framework was patented (US12330693B2). Fourth, a centralised computing architecture merging infotainment and ADAS on a single Snapdragon Ride Flex SoC was launched — described as the world’s first central vehicle computer of its kind — targeting a reduction in control unit count and cabling complexity.
“Computer program patents (1,075) now exceed any single sensor hardware category — a signal that Bosch’s competitive moat is increasingly built in software, not silicon.”
Explore Bosch’s full ADAS and sensor patent portfolio across all three phases in PatSnap Eureka.
Analyse Patents with PatSnap Eureka →Core Patent Landscape: Radar, LiDAR, Camera, and AI
The distribution of Bosch’s 11,745 patents across technology domains reveals a portfolio deliberately structured to cover every layer of the autonomous driving stack — from raw sensor data to AI inference to system integration. The single largest category is Computer Program (1,075 patents), which exceeds Radar (593), LiDAR (569), and Video Camera (410) individually.
Radar Technology Leadership
With 593 radar patents, Bosch holds a market-leading position in 77GHz and 24GHz radar across major automotive markets, as tracked by industry research published at Research and Markets. Key innovations include high-resolution imaging radar for object classification and contour mapping, interference mitigation using pseudo-random chirp sequences and frequency hopping for multi-radar coexistence, and the proprietary RF CMOS SoC with 22nm transistors that will enter serial production in 2026.
Camera and Vision Systems
The 410 video camera patents support a next-generation multi-purpose camera launching in China in 2026 with 8MP resolution and a 120° field of view, integrating with up to five radar units for L2-level functionality. A modular camera assembly patent covers electrically connected independent cameras to processor modules, reducing production costs and improving manufacturing flexibility. Thermal stability innovations maintain optical performance across a -40°C to +85°C operating range.
Bosch’s next-generation multi-purpose camera, launching in China in 2026, features 8MP resolution and a 120° field of view, and integrates with up to five radar units to deliver L2-level ADAS functionality.
Top Strategic Patents by Innovation Impact
Three patents stand out for their strategic value. US11587330B2 (filed 2019, granted 2023, estimated value $330,000) covers a visual analytics platform with coordinated visualisations — including size distribution, intersection-over-union, and area-under-curve metrics — enabling interactive identification of model weaknesses and accelerating AI training cycles. US12128890B2 (filed 2020, granted 2024, estimated value $280,000) covers closed-loop quality monitoring that adjusts sensor selection and measurement rates based on actual versus setpoint trajectory comparison, enabling energy-efficient autonomous driving. US12330693B2 (filed 2023, granted 2025, estimated value $46,000) covers the first feasible quantitative SOTIF failure-rate calculation method using driver status events and system fault severity.
Of Bosch’s 11,745 ADAS and sensor patents: 3,440 (29%) are active, 4,679 (40%) are pending, and 2,622 (22%) are inactive or abandoned. The 40% pending rate signals aggressive recent filing; the low abandonment rate reflects strong prosecution discipline in mature markets.
Current R&D Directions: Scalable ADAS and Edge Computing
Bosch’s 2024–2026 R&D agenda is organised around four interconnected priorities: a scalable three-tier ADAS product family, advanced localisation technology, edge computing and real-time processing, and robustness in edge cases. Each priority is backed by both patent filings and confirmed product launches.
Scalable ADAS Product Family
At Auto Shanghai 2025, Bosch launched a modular ADAS ecosystem spanning three tiers. The entry configuration uses camera-only hardware with optional radar for lane keeping, automatic emergency braking, and adaptive cruise control. The mid-tier combines multi-camera arrays with radar and centralised compute to enable hands-free highway driving and automatic lane changes — this variant entered production with a Chinese OEM in mid-2025. The high-end configuration deploys a full 360° video belt and complete sensor suite for urban autonomy including roundabout and junction handling, with a Summer 2025 launch. Six Chinese OEM orders have been secured, including confirmed partnerships with BAIC, Dongfeng, and Jetour.
Bosch secured six Chinese OEM orders for its scalable ADAS product family — including BAIC, Dongfeng, and Jetour — with mid-range variant serial production beginning with a Chinese OEM in mid-2025 and the high-end 360° perception variant launching in Summer 2025.
Edge Computing and Centralised Architecture
Bosch is targeting a reduction from 50 or more ECUs to three to five central vehicle computers, with a stated revenue target of €3 billion by 2026 from infotainment and ADAS computers alone, and a broader market opportunity of €32 billion by 2030. The centralised computing platform merges infotainment and ADAS on a single Snapdragon Ride Flex SoC. A 2024 patent covers vehicle-edge-cloud task allocation, distributing perception, planning, and control tasks across vehicle, edge, and cloud nodes to minimise latency.
Robustness in Edge Cases
Recent patents address scenarios that have historically challenged ADAS reliability. A LiDAR control patent covers beam intensity and point cloud density adjustment when detecting vulnerable road users such as pedestrians and cyclists. An automatic emergency braking patent covers fallback to radar-only configuration when camera performance degrades due to dim light or water spray. An ultrasonic sensor patent covers dynamic threshold adjustment for concealed sensors based on resonant frequency, compensating for temperature variation.
Map Bosch’s R&D priorities against competitor filings in real time using PatSnap Eureka’s company intelligence tools.
Explore Full Patent Data in PatSnap Eureka →Five Innovation Trends Shaping Bosch’s 2026 ADAS Roadmap
Five structural trends emerge from a cross-analysis of Bosch’s patent filings, product launches, and market positioning between 2022 and 2026. These trends are not independent — they reinforce one another in a coherent strategic direction.
Trend 1: From Hardware to Software Dominance
The 2010–2017 period was hardware-centric, focused on sensor design and signal processing. Between 2018 and 2021, the emphasis shifted to integration — sensor fusion and multi-modal perception. From 2022 onwards, the dominant theme is software-defined platforms: AI models, cloud integration, and over-the-air updates. The patent data confirms this: computer program patents (1,075) now exceed any individual sensor hardware category, and neural network patents (363) reflect deep learning adoption across the perception stack. Standards bodies including ISO have noted this broader industry shift toward software-defined vehicle architectures.
Trend 2: China-First Innovation Strategy
Bosch has adopted a localisation-first approach for ADAS, with China as the primary innovation hub. More than 50% of new Chinese cars were equipped with L2+ systems as of 2025. The mid-range ADAS variant entered production with a Chinese OEM in mid-2025, and navigation-linked assistance has been tailored for Chinese urban traffic patterns. This positions China not merely as a sales market but as a lead market for product development, with innovations then adapted for global deployment.
Trend 3: Quantitative Safety-by-Design
Bosch has pioneered quantitative SOTIF analysis, developing a formula-based failure rate calculation: λ = risk_factor × (RFIM_TTI / 3600s). This enables adaptive HMI process switching and reliability tuning when failure rates exceed safety thresholds, and addresses both ISO 26262 and ISO/PAS 21448. According to ISO, SOTIF compliance is increasingly required for Level 3 and above type approval in multiple jurisdictions.
Trend 4: Dynamic Sensor Suite Optimisation
Static sensor redundancy is being replaced by dynamic resource management. Real-time sensor activation and deactivation based on quality metrics — including trajectory deviation and object distance — allows measurement rate adjustment to balance precision and computational load. Patent US12128890B2 demonstrates energy efficiency gains through adaptive sensor operation. Context-aware fusion dynamically adjusts sensor weights based on environmental conditions including weather and traffic density.
Trend 5: Modular, Cost-Effective Architecture
A single hardware platform now spans entry-level to premium segments. Camera-only configurations serve cost-sensitive markets; radar-camera-LiDAR suites serve high-end autonomy. The modular camera assembly patent reduces production costs and improves manufacturing flexibility. This democratisation of ADAS is consistent with the regulatory direction set by the European Union, which is mandating specific ADAS features in all new vehicles by 2026, as noted by market research published at Research and Markets.
Emerging R&D Themes (2024–2026)
Three emerging patent clusters signal where Bosch’s next wave of innovation is headed. Ultra-wideband in-vehicle sensing patents cover occupancy detection and biometric data collection. Three-dimensional ultrasonic positioning patents cover MEMS sensor arrays with multi-directional transmission for accurate object height measurement. Generative AI for simulation patents cover point cloud generation using Doppler-enhanced training data — integrating Doppler velocity information into generative machine learning models for improved point cloud similarity calculation. Research published by IEEE has highlighted Doppler-enhanced radar simulation as a key frontier for autonomous driving perception robustness.
Market Outlook and Strategic Implications for Stakeholders
The ADAS chip market is expected to grow at a CAGR of 12.0% between 2021 and 2027, according to market analysis tracked by Yahoo Finance. Bosch’s stated target is a €32 billion market for infotainment and ADAS vehicle computers by 2030, with an interim milestone of €3 billion revenue from this segment by 2026. The EU’s mandate for specific ADAS features in all new vehicles by 2026 provides a regulatory tailwind for the entire supplier ecosystem.
Near-Term Milestones (2025–2026)
Four confirmed milestones define the near-term roadmap. Mid-2025 marks serial production of the mid-range ADAS variant with a Chinese OEM. Summer 2025 marks the launch of the high-end variant with 360° perception. 2026 marks both the launch of 200mm SiC wafer production in Roseville, California — backed by a $1.9 billion investment — and the deployment of the 8MP multi-purpose camera in China.
Competitive Advantages and Risk Factors
Bosch’s end-to-end ownership — from silicon (SoC, SiC) to software (AI models) to system integration — is a genuine differentiator: it is the only Tier 1 supplier manufacturing a complete radar system in-house. The modular architecture enables rapid time-to-market across segments, and the quantitative SOTIF methodology addresses Level 3+ regulatory requirements that competitors have not yet publicly solved.
Risk factors are also present. Bosch holds 569 LiDAR patents, but recent product emphasis on camera-radar configurations creates a potential vulnerability if solid-state LiDAR becomes cost-competitive. The centralised computing platform relies heavily on the Snapdragon Ride Flex SoC; the 2026 SiC production launch mitigates but does not eliminate chip supply dependency. Competition for AI and machine learning talent from technology companies including NVIDIA, Qualcomm, and Mobileye represents a sustained pressure on software leadership. These competitive dynamics are tracked by organisations including WIPO in their annual technology trend reports on autonomous vehicle innovation.
“Bosch is targeting a €32 billion market for infotainment and ADAS vehicle computers by 2030 — a figure that underscores why the software patent count now eclipses every sensor hardware category.”