The Rise of ToF Sensors: The New Favorite in IoT

Since last year, many sensor manufacturers and smartphone companies have turned their attention to ToF sensors. Even this year, companies like Infineon, AMS, as well as Apple, Huawei, and Samsung continue to promote the technology and application upgrades of ToF sensors, indicating that ToF sensors are not only gaining popularity but have indeed arrived.

Although major manufacturers are involved, from a technical perspective, ToF is not a particularly advanced technology. Simply put, it emits modulated light pulses through an infrared emitter, which reflects off an object and is received by a receiver, calculating the distance based on the round-trip time of the light pulses. Based on this principle, the early applications of ToF technology were relatively simple, primarily for distance measurement.

However, as the applications of ToF technology continue to expand, ToF sensors have entered the public eye mainly in the smartphone and tablet fields, focusing primarily on 3D ToF image sensors. Given that ToF sensors are currently most commonly used in imaging, this article will only discuss ToF image sensors.

Regardless of the smartphone manufacturer, the latest models are equipped with at least two cameras, demonstrating significant effort in camera quantity, which seems to be one of the few remaining innovations manufacturers can implement in smartphones. Many users believe that multiple cameras are merely a way for manufacturers to enhance photography performance, but the reality is different.

As ToF sensors gradually become standard in smartphones, the purpose of multiple cameras becomes clearer, allowing for various scene recognition applications, such as gesture recognition or 3D facial recognition for secure payments using front and rear lenses, as well as AR/VR applications that utilize ToF for 3D perception.

According to Figure 1, data from IHS Markit indicates that in 2019, the market size of ToF sensors in the 3D optical market exceeded $500 million, and their market share continues to grow. Although ToF is a latecomer compared to binocular and structured light solutions, it is likely just a matter of time before it achieves parity or even surpasses them.

Figure 1

Figure 2

From Figure 2, we can see that the current market share of ToF sensors in niche areas is primarily dominated by consumer electronics and automotive sectors.

However, we note that ToF image sensors not only have significant application prospects in consumer electronics but also have great potential in the IoT field. Following the smartphone and tablet markets, the next areas for ToF sensor market share include building detection, smart home, automotive central control, and drones.

With such a wide range of application scenarios, ToF image sensors benefit from their advantages over structured light and binocular RGB: they can quickly and accurately calculate depth information of objects in real time, and depth calculations are not affected by the surface gray characteristics of objects, maintaining centimeter-level accuracy regardless of distance, making them particularly suitable for applications with large range variations and high speed.

Additionally, since it is an active light source, the theoretical maximum detection distance of ToF sensors can reach 100m without harming human eyes, and the light source can be flexibly adjusted to meet distance requirements. Furthermore, ToF sensors exhibit significant advantages in anti-interference capability and cost.

Based on the aforementioned advantages of ToF sensors, I believe that they can be widely applied in the IoT field and are better suited for the fragmented scene requirements of IoT, offering greater flexibility.

Overall, for 2D recognition technologies that previously lacked depth information, ToF sensors can significantly increase recognition dimensions, enhancing safety, comprehensiveness, and accuracy of recognition.

Although ToF sensors exhibit good application scenarios in the IoT field, the market has not yet fully opened up. I believe the biggest obstacle is the disadvantages of ToF sensors in power consumption and resolution. Due to the use of active light sources, ToF sensors currently do not meet the low power consumption requirements of IoT hardware.

Moreover, limited by depth information capture, the more points the ToF solution needs to project, the better. Currently, ToF detectors can measure approximately 100 million measurement values per second, which limits the resolution of existing ToF systems to the centimeter level.

Therefore, addressing the multi-scene application issues of ToF technology itself, power consumption and resolution are unavoidable obstacles. However, a shift in thinking may present a breakthrough.

From the perspective of resolution and power consumption, integrating AI seems to be a possible method to overcome the bottlenecks of ToF technology. Since the imaging resolution of 3D ToF is far inferior to existing RGB binocular technology, AI algorithms can be used to synchronize RGB and ToF imaging information, narrowing the resolution gap required for 2D+3D image synchronization and compensating for the lack of depth information at object edges in ToF.

Currently, significant progress has been made in the development of AI chips or NPUs in China. Using both in conjunction with ToF instead of a general-purpose main processor can significantly reduce overall power consumption and further lower costs. If AI and ToF are integrated in research and development, it will help ToF further break through technological bottlenecks and extend more application scenarios.

Of course, for IoT, low power consumption is a necessity, but the demand for resolution may be more flexible. Many ToF sensor manufacturers are launching high-resolution ToF solutions at high costs, but whether downstream module and terminal manufacturers can overlook the high-cost factor is worth considering. Therefore, ToF sensors are still in the early stages of practical application in IoT, and related issues will exist, but in the field of 3D recognition, ToF will become one of the mainstream solutions due to its excellent features.