TE, one of the world's sensor and connector giants: How to Evolve Autopilot Sensing Technology

Hundreds of billions of connectors and sensors are embedded in large-scale networked physical devices, continuously collecting, aggregating and transporting massive data. Whether it is an advanced medical and health equipment that remotely monitors heartbeat rate and medication consumption, or an automated pipeline that can real-time visualize production and operation data, warn of faults in advance, or a smart home system that can be activated and shut down by mobile phones, as well as a burgeoning autopilot/unmanned driver, there are connections and sensing in all places where there is power, signal and data flow. Small sensors have spawned Big business. The goal of all sensors is very simple: to collect data (vibration, temperature, pressure, voltage, humidity) that can be analyzed by the algorithm for better real-time decision-making. The more data collected, the richer, the better the real-time analysis effect. Compared with dazzling terminal products, connectors and sensor companies upstream in the industrial chain are usually unknown, but they are invisible heroes. TE Connectivity, one of the world's largest manufacturers of sensors and connectors, is one of them, with sales of $14 billion in FY2018.

87% of TE revenue in FY2018 comes from applications in harsh environments. Full-load and high-speed factory assembly lines, deep ocean with destructive pressure, minimally invasive surgery, high-speed and high-risk racing events, the more this kind of rigorous environment can not be lost, the more TE appetite, the two Mars probes "Opportunity" and "Curiosity" in the United States have been equipped with TE temperature sensors, successfully landed on Mars.

"For example, to achieve 100Mbit/s or even 1Gbit/s transmission rate in high-speed automotive networks, which requires chipsets to achieve fast data transmission, requires a lot of sophisticated cable and connection technology. TE cable and connection solutions can help the system achieve such a high data transmission rate. Alan Amici, the company's vice president and chief technology officer of transportation solutions, told reporters, including Surging News, at a recent TE Chief Technology Officer Round Table in Shanghai.

Sensors were originally developed for large and expensive industrial platforms such as jet engines. These electronic "eyes" and "ears" help to identify defective parts and avoid failure in real work. As the cost decreases, sensor applications extend to all walks of life.

It is understood that TE's connection and sensor solutions are widely used in the fields of driverless vehicles, electric vehicles, next generation commercial aircraft, intelligent factories, high-speed data centers and advanced medical equipment. The company was named one of 63 "world-changing companies" by Fortune magazine in 2018.

The development of autopilot technology requires higher precision and synergy of high-efficiency sensors, antennas and data connection technology. The number of connected sensor components used in automobiles is increasing, and the amount of data transmitted is also increasing. In the increasing data traffic, large data packets requiring high bandwidth continue to occupy a large share.

Auto/Unmanned Vehicles rely mainly on light radar sensors, which emit lasers to detect nearby objects to determine distances, and then draw the vehicle's surroundings in 3D form. On this basis, the unmanned driving system can analyze data, issue instructions and execute them through the algorithm. As the first procedure, the accuracy of sensors directly affects the subsequent decisions and instructions whether they match the real environment.

In the process of automatic driving, wireless communication enables vehicles to keep in touch with each other, and can monitor the surrounding environment including pedestrians and cyclists, thus improving safety. On V2V (vehicle-to-vehicle) network, cars transmit information to each other, telling each other what they are doing, including speed, position, driving direction, braking, etc.

"Unmanned driving will use a lot of sensors and require a lot of data, such as radar, antenna and other on-board systems will involve a lot of data transmission. In addition, besides the automobile system itself, there are many information, including map information and V2V information, stored in the cloud. The key problem is how to realize the transmission of these large amounts of data. According to Alan Amici, automotive development is increasingly inclined to build high-speed data networks within automobiles. Previously, the local connection network of automobiles was used. Now it has been converted to the connection network based on ethernet, and the speed has been greatly improved. TE develops an automobile data network based on ethernet. Its data transmission rate can reach 100M bit/sec or 1G bit/sec. TE is developing the next generation of data networks that can deliver faster data rates.

In addition, TE is developing the next generation of high stability data connectors for critical security applications such as lidars and radars that generate huge amounts of uncompressed data.

In the past, some static manufacturing has gradually turned to more flexible flexible manufacturing, and robots have been put into large-scale application. In the future, the density of sensor layout in factories will be higher and higher. Industrial sensors not only require strict performance indicators, but also have a variety of types. Taking industrial robots as an example, the types of sensors involved include: three-dimensional vision sensor, force and torque sensor, collision detection sensor, safety sensor, seam tracking sensor, tactile sensor and so on.

David Brown, TE Vice President and Chief Technical Officer of Industrial Solutions, believes that future smart factories with ubiquitous sensing technology have a large amount of data that can be used by AI to help factories make "informed decisions" and make daily maintenance of factories more accurate. "In terms of R&D, we think there are two main trends. The first is that after more and more sensor technologies are applied in factories, our connectors and sensors will be deployed faster. In addition, through many computer images and continuous improvement of sensing technology, the scale of data will be greatly increased.