Modern vehicles have cameras, radar, LiDAR, and increasingly capable driver-assistance software. Yet one basic limitation remains: a sensor cannot detect what it cannot see.
A truck can hide a pedestrian at an intersection. A building can block an approaching vehicle. Dense traffic can conceal a cyclist or road hazard until there is very little time to react.
Vehicle-to-Everything, or V2X, technology is starting to address this problem by giving vehicles access to information collected beyond their own sensors.
Instead of relying only on what one car sees, connected systems can use observations from nearby vehicles, roadside equipment, and communication networks. Recent developments highlighted in the June 2026 report combine shared sensing, smarter data exchange, and integrated sensing and communication, or ISAC, to improve awareness when a vehicle’s direct view is blocked.
Four recent patent applications show how this field is developing. Sony is working on how vehicles decide which sensing information to trust. Zhejiang University combines observations from several vehicles into a shared road view. Southeast University reduces the amount of visual data that needs to move through the network. Dalian University applies AI to the infrastructure managing sensing and communication.
Taken together, these patents show an important shift: vehicle perception is expanding from an onboard function into a connected system.
Why Better Sensors Alone Do Not Solve Blind Spots
Most advanced driver-assistance systems start with sensors mounted on the vehicle. These work well when an object is visible, but real roads constantly create blocked views.
The report illustrates a common example: a large vehicle prevents one car from seeing a pedestrian near an intersection, while another connected vehicle or roadside system has a clearer view and can share that information through V2X.
This changes the engineering problem. Improving cameras or radar is no longer the only route to better road awareness. Vehicles also need to combine information coming from several viewpoints.
That creates harder questions. Which external source should the vehicle trust? How should observations of the same object be matched? How much information can be exchanged without slowing the network? And how should communication resources be managed when many vehicles need sensing information at the same time?
The four patents address different parts of this challenge.
Sony Is Focusing on Which Sensor Information Deserves More Trust
Sony’s CN121893975A, with a 2024 priority date, combines onboard vehicle sensing with environmental information received through V2X and ISAC.
The interesting part is how the system treats those inputs.
It does not give every source the same importance. Instead, the contribution of each source can change according to factors such as reliability, confidence, road conditions, and proximity to an event.
Imagine a vehicle driving through heavy rain. Its camera struggles to identify something near the road, while a roadside sensor has a clearer view. A confidence-aware system can place greater weight on the more reliable observation rather than treating both inputs equally.
For R&D teams, the message is important: adding more sensors and outside data does not automatically improve perception. The system also has to determine which information deserves trust at that moment.
For IP teams, this expands the area worth monitoring beyond cameras, radar, and LiDAR hardware. Patents covering confidence scoring, sensor weighting, reliability checks, and context-based data fusion become increasingly relevant as vehicles draw on more external information.
Zhejiang University Is Building a Shared View of the Road
Zhejiang University’s CN121697664A, with a November 2025 priority date, addresses blind zones created by dense traffic and large vehicles.
The patent allows neighboring vehicles to share perception information through V2X. Those observations are combined into a unified vector map, which can then support AI-based warnings and driving assistance.
Consider three vehicles travelling through the same intersection. One sees a pedestrian. Another detects an approaching vehicle. The third has its view blocked by a bus.
Instead of leaving the third vehicle with incomplete information, observations from the other vehicles can fill in parts of the road scene that it cannot directly observe.
The report describes this as a move from isolated vehicle sensing toward shared environmental intelligence.
The concept sounds simple, but the technical work behind it is not. Information from moving vehicles has to be aligned to the same road scene, kept current, and combined quickly enough to support a driving decision. The system also needs to deal with cases where two sources report different information.
For IP teams, this makes shared mapping, multi-vehicle sensing, data alignment, and collaborative warning systems important areas to follow alongside traditional V2V communication.
Southeast University Is Tackling the Data Bottleneck
Shared road awareness creates another problem: bandwidth.
Cameras and other sensors generate large amounts of information. If every roadside camera and connected vehicle continuously transmits raw data, communication delays and network congestion quickly become serious barriers.
Southeast University’s CN121697664A, with a December 2025 priority date, addresses this by reducing the amount of visual information that needs to be transmitted. The report identifies blocked views, different viewing angles, timing mismatches, and high communication overhead as major challenges for existing vehicle-road sensing systems.
The proposed approach extracts useful features from roadside-camera images, compresses selected information, and creates semantic masks that distinguish meaningful parts of the scene. That processed information is then combined with features from the vehicle’s own camera for 3D object detection.
In simpler terms, the system tries to send what matters in the image rather than the whole image.
This is a critical R&D problem because connected sensing does not scale simply by increasing network capacity. Systems also need to become better at deciding what information is worth transmitting.
The patent points to a wider innovation area around feature compression, semantic communication, edge processing, and selective data sharing.
Dalian University Is Bringing AI Into the Network Layer
The first three patents focus largely on perception and information exchange. Dalian University’s CN121940783A, with a January 2026 priority date, looks at the network supporting those activities.
The patent concerns integrated sensing and communication, or ISAC, where wireless infrastructure supports both communication and environmental sensing.
Managing such a network becomes difficult when many base stations, vehicles, sensing requests, and communication tasks compete for resources.
Dalian University’s approach treats different parts of the network as learning agents. Generative AI creates additional training examples, while the agents learn how to manage areas such as transmission power, beamforming, and base-station activity.
The broader aim is to make network decisions more responsive as traffic and sensing demands change, rather than relying only on heavy centralized optimization.
This matters because large-scale connected driving requires more than accurate sensors. The supporting infrastructure also has to decide where network capacity is needed and how sensing and communication demands should be balanced.
For R&D teams working in telecommunications, edge computing, or connected mobility, network intelligence is becoming part of the vehicle-perception problem rather than a separate infrastructure issue.
Four Patents, Four Parts of the Same Problem
The four patents address different technologies, but together they reveal a clear development pattern.
| Patent | Main problem being addressed |
| Sony – CN121893975A | Deciding how much trust to place in different sensing sources |
| Zhejiang University – CN121697664A | Combining observations from multiple vehicles |
| Southeast University – CN121697664A | Sharing useful perception data without overloading the network |
| Dalian University – CN121940783A | Managing sensing and communication resources intelligently |
The progression matters.
Vehicles first need information from more than one viewpoint. That information must then be combined correctly. The useful parts have to move through the network efficiently, and the infrastructure must manage the resources supporting those exchanges.
This is why V2X development is moving beyond basic messages such as location, speed, or collision warnings. The network is becoming part of how the vehicle builds its understanding of the road.
What R&D Teams Should Watch
For R&D teams, four areas stand out.
Confidence-aware data fusion addresses which sensor or external source should be trusted under changing conditions.
Shared road mapping expands vehicle awareness by combining observations from different viewpoints.
Semantic and compressed communication reduces the network load created by richer sensing data.
And AI-managed ISAC infrastructure connects road perception with communication-resource decisions.
These technologies solve different problems, but their direction is similar: connected driving systems are becoming less dependent on what a single vehicle can observe by itself.
What IP Teams Should Watch
For IP teams, the important lesson is that connected-vehicle innovation will not always appear under a simple “V2X” label.
Relevant patents can sit across sensor fusion, computer vision, shared mapping, wireless communication, feature compression, 3D detection, ISAC, AI-based network management, and vehicle-road collaboration.
That makes problem-based patent monitoring particularly useful.
Instead of searching only for V2X technologies, teams can track the underlying problems companies and researchers are trying to solve: blocked visibility, conflicting sensor information, shared mapping, communication overhead, latency, and network resource allocation.
This approach can reveal important developments even when inventors describe similar problems using very different technical language.
Want the Full Patent Landscape or Ongoing Updates?
The patents covered in this article represent only a snapshot of the innovation happening in this technology area.
A broader patent landscape can help your team understand who else is working on similar solutions, which technical approaches are gaining attention, where new filings are appearing, and which organizations may offer strategic opportunities.
If your R&D or IP team wants to go deeper, we can help identify:
- Emerging patents and technology directions
- Key companies, startups, universities, and inventors
- Competitor activity and portfolio movement
- Potential licensing opportunities
- Possible R&D or technology collaboration partners
- Companies or technologies that may be relevant for acquisition or strategic investment
- White spaces and underexplored technical areas
- New patent filings worth tracking over time
Whether you need a complete patent landscape or want to continuously track breakthrough patents, competitors, and new technology developments, tell us what your team is exploring.
Fill out the form below and let us know the technology, competitor, or opportunity you want to investigate.
The Vehicle Is No Longer the Only Viewpoint
The strongest signal from these four patents is not simply that vehicles are communicating more.
It is that communication is becoming part of perception.
Sony is exploring how outside sensing information should be weighted. Zhejiang University is building shared road views from several vehicles. Southeast University is making those exchanges lighter and more practical. Dalian University is working on the infrastructure needed to manage sensing and communication together.
For R&D teams, these patents reveal where technical problems are moving. For IP teams, they show innovation forming at the intersection of automotive sensing, AI, wireless networks, and road infrastructure.
The next step in connected driving is not simply giving a vehicle better sensors.
It is giving the vehicle access to viewpoints it does not have itself.
