The availability of low cost, solid state sensors in huge volumes will allow for nearly ubiquitous sensing of our entire world, as a part of the Internet of Things. From surveillance cameras for keeping property secure to vibration sensors detecting the next earthquake, we are poised to live in an era when we will be able to understand the physical environment on a scale we have never been able to before. Tiny electronic devices, embedded in everything from toasters to bridges will be possible. Is the tsunami of data this ocean of sensors will provide truly useful? And who will succeed in this space – those who can collect the most data, or those who understand best how to analyze it? Perhaps the sheer volume of information will hide the important signals in a flood of noise. Solving this problem will be difficult; there are however some key principles that can help in this regard – choosing the correct sensor for the task, and taking advantage of the principle of sensor fusion to increase the depth and quality of the information gathered.
We are moving from an era of sensors based on mechanical and electromechanical principles to an era of almost purely electronic solid state devices. This brings huge improvements in reliability, size and cost. These advances mean that sensors can be placed in many new places, and can be used for many new applications. Radar has also seen progress from these developments, and radar based systems are now common for uses such as automotive anti-collision sensing or motion based door opening. Many of these products are now relying on packaged silicon ICs assembled using conventional pick and place processes and PCB based antennas. This is in contrast to “traditional” microwave systems, which relied on discrete mechanical and electronic components, often hand assembled and tuned. Radar is thus poised to become another commodity sensing technique. Several tech industry players have started to unveil concepts using these technologies, such as Google with its Project Soli for user interfaces based on hand gestures. This project has developed a radar front end in a single plastic package, intended for integration in smartwatches and phones. Through this project, Google has also illustrated two of the inherent advantages of radar over competing sensing technologies – it is concealable and non-contact.
Sivers IMA has followed these developments in radar closely, and has developed several products taking advantage of these advances. The RS3410X/00 is one such product – an integrated radar front end operating at 9.5 to 9.975 GHz intended for safety and security applications. Due to the inclusion of a frequency synthesizer it can be used for FMCW based distance measurement, in addition to Doppler based motion and speed detection. It uses reliable and low cost packaged components, and is designed for surface mounting, enabling low cost, and integration with PCB antennas.
Since one can measure motion, speed or distance, a sensor using the RS3410X/00 can be used to detect a wide variety targets in the environment, either as a primary device, or in combination with another sensor to provide redundancy or data fusion. The inherent immunity to weather and light conditions provides an advantage over technologies such as video cameras, and laser or infrared technology. One can also reduce the likelihood of false alarms due to the type of information that can be extracted from a radar signal.
So how would all of these features and concepts combine in a real-life application to provide a better solution in this world covered in sensors? Say one wants to use image recognition techniques to measure and monitor traffic flow along a multi-lane road. In most conditions this would be a straightforward problem using a single camera, but sometimes special cases arise – perhaps it is snowing, reducing visibility. One could try to add more or closer cameras to compensate, but this illustrates the problem of a flood of the wrong information, as the signal processing problem now becomes far more complex and therefore expensive. Using a sensor based on the RS3410X/00 in combination with the video camera would allow this problem to be solved, as the radar can penetrate the precipitation, while not adding significant computing overhead. The radar would also enable accurate speed measurement of the vehicles, enabling a deeper level of understanding of the traffic being measured. All of these factors combine to make it extremely useful in solving two of the main problems we will encounter in this new world of ubiquitous sensing – selecting the correct sensor for the application, and understanding the data that is gathered.
Alex Vaivars
Product Manager – Radar