Periodic Reporting for period 1 - DeepSight (Detection of pedestrians and cyclist outside a line of sight)
Berichtszeitraum: 2022-06-01 bis 2023-11-30
The problem:
The number of road accidents continues to increase dramatically each year. Cyclists and pedestrians, being a part of the daily traffic, keep challenging drivers thus making safety issues increasingly important. According to the European Commission (Eurostat), light goods vehicles, mopeds, bicycles, motorcycles, and pedestrians are involved in more than 50% of road accidents. Improving driving safety, being always a subject to technological efforts, is one of the most important challenges of modern society.
Limitation of current solutions:
All the existing tools face the same fundamental limitation – they require a direct line of sight to an object.
The solution:
The vast majority of conventional solutions are based on sending an interrogating signal and listening for echoes from an object. Visible and infrared (IR) cameras detect ambient radiation, imaging a venue. The quest for a better angular and range resolution pushes the technologies towards using high-frequency electromagnetic waves, which can neither diffract nor penetrate through obstacles. Consequently, all those approaches are based on having an unperturbed line of sight. However, low-frequency waves are never used in automotive applications owing to the low range and angular resolutions they can grant, as it is commonly believed. While this statement is almost a ground truth in the field, it can be bypassed if preliminary information on an object does exist. This includes and not limited to rotating bicycle wheels, human breath, and several others. Here we develop a low-frequency radar for detecting cyclists and pedestrians with a resolution, sufficient to ensure road safety and assist drivers to make the right decisions. Our new system is based on (i) genetically designed antenna elements with reduced footprints and boosted efficiencies, (ii) generation and detection of smart radar signals, and (iii) real-time processing and decision-making algorithms.
The number of road accidents continues to increase dramatically each year. Cyclists and pedestrians, being a part of the daily traffic, keep challenging drivers thus making safety issues increasingly important. According to the European Commission (Eurostat), light goods vehicles, mopeds, bicycles, motorcycles, and pedestrians are involved in more than 50% of road accidents. Improving driving safety, being always a subject to technological efforts, is one of the most important challenges of modern society.
Limitation of current solutions:
All the existing tools face the same fundamental limitation – they require a direct line of sight to an object.
The solution:
The vast majority of conventional solutions are based on sending an interrogating signal and listening for echoes from an object. Visible and infrared (IR) cameras detect ambient radiation, imaging a venue. The quest for a better angular and range resolution pushes the technologies towards using high-frequency electromagnetic waves, which can neither diffract nor penetrate through obstacles. Consequently, all those approaches are based on having an unperturbed line of sight. However, low-frequency waves are never used in automotive applications owing to the low range and angular resolutions they can grant, as it is commonly believed. While this statement is almost a ground truth in the field, it can be bypassed if preliminary information on an object does exist. This includes and not limited to rotating bicycle wheels, human breath, and several others. Here we develop a low-frequency radar for detecting cyclists and pedestrians with a resolution, sufficient to ensure road safety and assist drivers to make the right decisions. Our new system is based on (i) genetically designed antenna elements with reduced footprints and boosted efficiencies, (ii) generation and detection of smart radar signals, and (iii) real-time processing and decision-making algorithms.
Radar waveforms and efficient postprocessing: fast method to explore very complex scenarios to optimize radar waveforms
Identify the level of signal reflected/transmitted to the target: Non-line of sight detection is possible in the presence of metal obstacles
Asses outdoor detection capabilities: Moving objects can be identified from the clutter
Compact antennas: The smallest (new state-of-the-art) supergain antenna demonstrated
A fast method to steer antenna pattern: A novel method for fast steering is achieved – potentially, the MHz rate reported
Triangulation on a non-line-of-sight object via its micro-Doppler signature: subwavelength resolution in localization has been demonstrated
Identify the level of signal reflected/transmitted to the target: Non-line of sight detection is possible in the presence of metal obstacles
Asses outdoor detection capabilities: Moving objects can be identified from the clutter
Compact antennas: The smallest (new state-of-the-art) supergain antenna demonstrated
A fast method to steer antenna pattern: A novel method for fast steering is achieved – potentially, the MHz rate reported
Triangulation on a non-line-of-sight object via its micro-Doppler signature: subwavelength resolution in localization has been demonstrated
New Approach for Optimizing Radar Signals
The task of radar design is traditionally application-specific, both in terms of the signal processing algorithms as well as the hardware implementation. Only after overcoming the challenges of fabrication and validation can the final design be tested in the field. Assessing performances is made at numerous locations and complicated scenarios, involving multiple cooperating targets, various clutter environments, and other aspects. Today, the data collected from such field experiments cannot be straightforwardly used for designing and testing other radar systems, forcing every new design to undergo the same costly evaluation in the same real-world scenarios. Here we demonstrate the concept of a universal radar that can emulate any other radar system as if it were itself active in the investigated scene. The method is based on post-processing the frequency response measured by a broadband stepped frequency continuous wave radar, such as a vector network analyzer. This architecture achieves extremely low receiver noise figures with ultra-wide synthetic bandwidth and high transmitting power while enabling efficient frequency spectrum sharing for multiple nearby operating systems. This approach not only provides an all-purpose radar toolkit for different applications but more crucially enables the acquired data from the interrogated site to be used for the design and analysis of almost any other radar system, significantly cutting the time and cost of new developments. The method suggests the potential monetization of raw data across diversely different radar platforms.
The Smallest Supergain Superbandwidth Antenna
High-gain antennas are essential hardware devices, powering numerous daily applications, including distant point-to-point communications, safety radars, and many others. While a common approach to elevate gain is to enlarge an antenna aperture, highly resonant subwavelength structures can potentially grant high gain performances. The Chu-Harrington limit is a standard criterion to assess electrically small structures and those surpassing it are called superdirective. Supergain is obtained in a case when internal losses are mitigated, and an antenna is matched to radiation, though typically in a very narrow frequency band. Here we develop a concept of a spectrally overlapping resonant cascading, where tailored multipole hierarchy grants both high gain and sufficient operational bandwidth. Our architecture is based on a near-field coupled wire bundle. Genetic optimization, constraining both gain and bandwidth, is applied on a 24-dimensional space and predicts 8.81 dBi realized gain within a half-wavelength in a cube volume. The experimental gain is 6.158.22 dBi with 13% fractional bandwidth, which is the best performance in the field of superdirective antennas now. The developed approach can be applied to low-frequency (e.g. kHz-MHz) applications, where miniaturization of wireless devices is highly demanded.
Electro-optical Antenna Steering - high rate low-cost beamforming
The ability to obtain dynamic control over an antenna radiation pattern is one of the main functions, desired in a vast range of applications, including wireless communications, radars, and many others. Widely used approaches include mechanical scanning with antenna apertures and phase switching in arrays. Both of those realizations have severe limitations, related to scanning speeds and implementation costs. Here we demonstrate a solution, where the antenna pattern is switched with optical signals. The system encompasses an active element, surrounded by a set of cylindrically arranged passive dipolar directors, functionalized with tunable impedances. The control circuit is realized as a bipolar transistor, driven by a photodiode. Light illumination in this case serves as a trigger, capable of either closing or opening the transistor, switching the impedance between two values. Following this approach, a compact half-a-wavelength footprint antenna, capable of switching between 6 dBi directional patterns within a few milliseconds’ latency was demonstrated. The developed light activation approach allows constructing devices with multiple almost non-interacting degrees of freedom, as a branched feeding network is not required. The capability of flexible switching between multiple electromagnetic degrees of freedom opens pathways to new wireless applications, where fast beam steering and beamforming performances are required.
The task of radar design is traditionally application-specific, both in terms of the signal processing algorithms as well as the hardware implementation. Only after overcoming the challenges of fabrication and validation can the final design be tested in the field. Assessing performances is made at numerous locations and complicated scenarios, involving multiple cooperating targets, various clutter environments, and other aspects. Today, the data collected from such field experiments cannot be straightforwardly used for designing and testing other radar systems, forcing every new design to undergo the same costly evaluation in the same real-world scenarios. Here we demonstrate the concept of a universal radar that can emulate any other radar system as if it were itself active in the investigated scene. The method is based on post-processing the frequency response measured by a broadband stepped frequency continuous wave radar, such as a vector network analyzer. This architecture achieves extremely low receiver noise figures with ultra-wide synthetic bandwidth and high transmitting power while enabling efficient frequency spectrum sharing for multiple nearby operating systems. This approach not only provides an all-purpose radar toolkit for different applications but more crucially enables the acquired data from the interrogated site to be used for the design and analysis of almost any other radar system, significantly cutting the time and cost of new developments. The method suggests the potential monetization of raw data across diversely different radar platforms.
The Smallest Supergain Superbandwidth Antenna
High-gain antennas are essential hardware devices, powering numerous daily applications, including distant point-to-point communications, safety radars, and many others. While a common approach to elevate gain is to enlarge an antenna aperture, highly resonant subwavelength structures can potentially grant high gain performances. The Chu-Harrington limit is a standard criterion to assess electrically small structures and those surpassing it are called superdirective. Supergain is obtained in a case when internal losses are mitigated, and an antenna is matched to radiation, though typically in a very narrow frequency band. Here we develop a concept of a spectrally overlapping resonant cascading, where tailored multipole hierarchy grants both high gain and sufficient operational bandwidth. Our architecture is based on a near-field coupled wire bundle. Genetic optimization, constraining both gain and bandwidth, is applied on a 24-dimensional space and predicts 8.81 dBi realized gain within a half-wavelength in a cube volume. The experimental gain is 6.158.22 dBi with 13% fractional bandwidth, which is the best performance in the field of superdirective antennas now. The developed approach can be applied to low-frequency (e.g. kHz-MHz) applications, where miniaturization of wireless devices is highly demanded.
Electro-optical Antenna Steering - high rate low-cost beamforming
The ability to obtain dynamic control over an antenna radiation pattern is one of the main functions, desired in a vast range of applications, including wireless communications, radars, and many others. Widely used approaches include mechanical scanning with antenna apertures and phase switching in arrays. Both of those realizations have severe limitations, related to scanning speeds and implementation costs. Here we demonstrate a solution, where the antenna pattern is switched with optical signals. The system encompasses an active element, surrounded by a set of cylindrically arranged passive dipolar directors, functionalized with tunable impedances. The control circuit is realized as a bipolar transistor, driven by a photodiode. Light illumination in this case serves as a trigger, capable of either closing or opening the transistor, switching the impedance between two values. Following this approach, a compact half-a-wavelength footprint antenna, capable of switching between 6 dBi directional patterns within a few milliseconds’ latency was demonstrated. The developed light activation approach allows constructing devices with multiple almost non-interacting degrees of freedom, as a branched feeding network is not required. The capability of flexible switching between multiple electromagnetic degrees of freedom opens pathways to new wireless applications, where fast beam steering and beamforming performances are required.