Traditional human-computer interface is accomplished by means of input devices such as keyboard and mouse. Recent studies have shown that increasingly, hand/finger gesture recognition is becoming attractive in many applications including human-computer interface, and KaiKuTeK is committed to this development using mmWave radar.

Using matured CMOS technology, KaiKuTeK has successfully integrated millimeter wave radar, antenna, AI accelerator, DSP, and microprocessor into a single chip running Machine Learning algorithm.

Due to this high level of integration, K60168 has a simplified design flow with not only shortened design cycle but also reduced overall BOM cost. Our radar sensor’s high resolution and permeability broaden the range of application which traditional sensor technologies cannot reach. Most sensor solutions in the market do not offer built-in AI accelerator. Data collected by the sensor will need to be sent to high performance CPU/ MCU or even the cloud to do the necessary computing, especially when immediate response is required as an interface device.

Apart from wearable devices, in this post-COVID era, people tend not to touch the public information station to protect themselves. For this reason, air gesture control becomes an attractive alternative to existing touch screen sensor. In addition, gesture control brings innovative, intuitive user experience to many applications by using KaiKuTeK's 60GHz mmWave Radar sensor. The K60168 also provides rawdata output mode, allowing customers to develop their own algirothm based on different application scenarios and domain knowledge.

Solutions for Gesture Recognition


  • K60168-P Product Brief V1.01