Shulou(Shulou.com)06/03 Report--

We have talked a lot about smart cities. More and more technology companies have joined this industry proposition, from self-driving, car-road coordination, and then to the urban brain, a series of software and hardware elements are booming, vowing to make the traffic life look new.

By the end of 2018, there are 1167 intelligent transportation projects (excluding highway informatization) in China.

However, at a time when people are generally looking forward to the story of cars and roads, many "indications" of the integration period have also begun to emerge. Driverless car accidents are common, the cloud brain is under great pressure in front of the huge amount of data, and the costs of research and development and infrastructure remain high, but the efficiency of urban traffic has not improved proportionally.

If intelligent transportation is compared to a good play in the future of the city, then the preliminary preparatory work has been completed. The core team is nothing more than cloud brains, chips, sensors, cameras, concubines are tired of talking. As for what kind of work will be presented to the public in the end, it depends on how the director makes the characters pick up the correct script and offer the best acting skills.

So we talked to Westerners who specialize in high-end sensors, MEMS chips and systems to discuss what roles sensors play in the Smart City blockbuster.

(note: Xi Ren Ma Group is a leader in MEMS chip and sensor technology, with all-round capabilities such as MEMS chip and sensor design, manufacturing, testing, material synthesis, and strong independent core technical strength.)

The data-woven "urban neural network" hides several important travel bug

Before the sensor starts its performance, we need to help the city's neural network "catch bugs" and find out the size of the bug that affects perception and decision-making.

In the process of urban digital infrastructure, a large number of sensors and systems land every day. But there are several obstacles to these terminal data distances that really serve the urban brain:

1. Integrity of data collection. Modern cities and people's travel trajectories often revolve around complex and close traffic networks. On the one hand, this requires the urban system to cover every corner and form a complete and huge "neural network"; on the other hand, the high-speed flow of cars and people also challenges the real-time and accuracy of data acquisition. If there is a "blind spot" in the data, then no matter how powerful the urban brain is, traffic congestion, ground safety and other problems are still a "dead end".

two。 Compatibility of data collaboration. Another problem with urban data is the multi-variety and multi-mode of data sources. For example, only to solve the problem of traffic jam, we need road traffic data, vehicle GPS data, weather forecast, road condition health monitoring data and so on. Only by merging different kinds of sensors and kneading various data sources together, can we realize the real-time grasp and control of traffic conditions and actively influence intelligent decision-making.

3. High efficiency of data processing. At present, the data collection programs in many cities are realized through cameras and other means. This leads to two problems: first, the visual data of the camera needs a series of processing, such as semantic analysis, feature extraction, image understanding and so on, which puts forward higher requirements for computing resources, hardware configuration, data storage and so on. Second, large-scale real-time high-precision visual data are sent to the cloud, which can easily cause data accumulation and processing delay, become a heavy burden on the "urban brain" and delay some needs that require real-time response.

In the final analysis, the intelligence of a city should not only rest on the surface of being digitalized, but should really enable every inch of texture of the city to perceive and apply data, so that every "nerve impulse" caused by any ground event can reach the "brain center" of the city quickly and smoothly.

Turning Urban roads into Neural Networks: the brain and Logic of Western Man and Horse

With regard to the construction of urban data system, there is no "basic Law" at present, and the participants are all groping for themselves. Among them, the intelligent transportation system of the West has played a role in the construction of "neural networks" in many cities.

It is obvious that the data problem in the field of urban transportation can not be solved by a single software and hardware. As a result, the West Renren Horse Shenxing Network has also broken down the data network into three steps:

The first step: the comprehensive laying of multimodal sensors

Step 2: a comprehensive solution for multi-module monitoring

The third step: the real intelligent traffic with the linkage of rickshaws, people and roads.

The first is the data-aware network. As we mentioned earlier, the most prominent problem of urban intelligent transportation is the single way and type of collection, resulting in insufficient dimension and accuracy of the data, unable to take care of the huge detail corners.

Based on this, the west horse intelligent transportation system designs the traffic neural network from the top design to the terminal layout. Reflected in the specific hardware deployment, it is to combine multi-mode sensors with all kinds of data acquisition modules to comprehensively transform urban roads according to local conditions.

For example, through the street lamp sensor, the people and objects on the road can be located with high precision by 10cm. By setting magnetic coordinates on the ground, the movement of each 10cm of the vehicle can be well known. In the complex structural environment such as tunnel, ultrasonic rangefinder, static level, crack sensor, vibrating string collector and other sensors are used to do comprehensive work, and combined with vehicle networking, the data needed for traffic travel are collected in a deep and multi-modal system, so that every "nerve impulse" of external stimulation can be sensed by the city in a three-dimensional way.

Then there is the scene-based multi-module monitoring scheme.

This part is a bit like the brain partition of the neural network, the data perception network and the processing module together constitute the monitoring module of the vertical scene, providing clear and specific capabilities.

At present, the west horse intelligent transportation system can provide users with different modules, such as positioning and navigation, communication transmission, climate monitoring, bridge monitoring, tunnel monitoring, geological hazard monitoring and so on.

Taking the ground disaster module as an example, the West Ren Horse Intelligent Transportation system deploys sensors, data acquisition, data transmission and other subsystems on the slope and other road environment to monitor natural disasters in real time. These data will be uploaded to the database and processing control center in real time. Once an abnormal situation occurs, the safety evaluation can be carried out according to the actual situation, and the early warning subsystem will issue a hierarchical early warning upward, so as to achieve the function of monitoring slope safety and eliminating hidden dangers in time.

Finally, it is necessary to link different "sub-modules" such as roads, cars and people to form a complete urban digital system with clear details, multiple functions and rich scenes. Then with the help of the majestic computing power of the supercomputer, the intelligent management of urban traffic is completed.

Take the most common "urban disease"-traffic congestion, for example, once roads, infrastructure, people and cars are connected by the urban neural network, once the surrounding vehicles change, the induction between cars will be turned on in real time. The changes of the external environment, such as special climate, network disconnection and abnormal traffic flow, can be sensed and processed by the ubiquitous sensor network in time, and then transmit the instructions to the car owners by the urban brain to avoid obstacles reasonably. Such a non-corner urban network, naturally will not be easy to "myocardial infarction".

In this way, from synapses (sensor systems), to independent brain regions (vertical modules), to urban brains (intelligent transportation systems), the intelligent transportation system of the West Horse is through this interlocking sketch. Let the real intelligent city come to us earlier.

The role-playing of sensors on the road of urban intelligence

So far, data, intelligence and the city have completed a deep combination of software and hardware integration. So, we might as well briefly summarize what roles sensor systems should play in the construction of smart cities.

1. Planner

In the details of the deployment of the intelligent transportation system, we will find that the sensor system is not only carrying out road reconstruction work by machinery, but also fully combining with the practical problems in the development of urban traffic, through digital and intelligent means, to deal with complex and changeable traffic conditions.

In other words, a good data planner can make urban life real, and then improve the ability of urban services from the overall situation.

two。 Guardian

Urban digitization has attracted countless manufacturers to get a piece of the pie, but chip manufacturers, infrastructure enterprises, data maintenance and other separate ways, the resulting data fault, coordination imbalance, is the "hidden killer" lurking in the urban neural network.

On the other hand, the most basic sensor network can integrate car and car, car and person, car and network from the broadest data dimension.

Just like the west horse intelligent transportation system, under the premise of bad environment, it can do work through signal state, weather monitoring, tunnel health monitoring, centimeter magnetic grid and so on. To solve the communication, positioning, navigation and many other problems that vehicles may encounter.

It is such a unified data system that is so detailed to the level of "capillaries" that driving safety can be ensured to the maximum extent.

3. Deep tiller

In the real traffic network, there are many vertical and subdivided small environments, and each structure has its own special deployment requirements, which is not estimated enough to take care of one or the other in the transformation of road software and hardware. Let the vehicle on the road can not walk far, can not walk fast.

In this regard, the West has made careful and meticulous arrangements for special highway scenes such as bridges, tunnels and slope disasters. Based on this, autopilot, vehicle-road coordination and other technologies can really go out of the proving ground and let everyone see the real value.

At present, it seems that the city must attack on both sides with a comprehensive unified layout and vertical landing in order to truly realize the awakening of wisdom.

Therefore, the comprehensive deployment and cooperative operation of multi-mode and multi-kinds of sensors become very important. Otherwise, many nodes of urban roads can only be like hemiplegic patients, no matter how the body moves, it is difficult to be accurately perceived by the brain.

Perhaps in the future, when we review the process of intelligence in Chinese cities, we will find that when terminal founders such as Westerners and horses throw themselves into it, they chisel out "neurons" and build digital pipelines. it is only then that wisdom begins to flow, transmit and collide in the city, and finally sprout unprecedented wisdom.

Only when every key character takes seriously his own script of the times and the proudest wisdom drama of urban civilization can it really begin.

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.