Seeing objects hidden behind a concrete wall or detecting someone before they appear around a corner seems like a skill straight out of comics, but researchers at the University of Pennsylvania transformed that idea into real technology, developing holoradar, a system that combines radio waves and artificial intelligence to allow robots and autonomous vehicles to perceive what is beyond the field of vision.

The most impressive thing is that this ability does not depend on cameras or lighting, it uses the structure of the environment itself as part of the perception system; today any system without this new technology basically depends on what is directly visible. A pedestrian hiding behind a truck or a forklift appearing behind a shelf represents dangerous blind spots.

The solution found by the researchers was to convert a characteristic of radio waves previously considered a limitation and an extraordinary advantage, Because radio waves have much longer wavelengths than light, this means that apparently rough surfaces such as walls, floors and ceilings begin to behave like true electromagnetic mirrors.

The holoradar sends signals that are reflected on these surfaces, go around obstacles, reach hidden objects and return to the sensors, but interpreting this network of reflected signals is extremely complex, this is where artificial intelligence comes in. Algorithms developed by the team filter out all interference, analyze signal patterns and reconstruct a three-dimensional image of the hidden environment. This makes it possible to identify, for example, a pedestrian approaching an intersection before he or she is visible to the vehicle's cameras.

This strategy offers a huge advantage over traditional optical systems. While laser sensor cameras can suffer from darkness, smoke, fog or sudden changes in lighting, holoradar continues to blend virtually unaffected by ambient conditions and these discrete, unhyped technologies are shaping the future of humanity when applied to other innovations we use in our daily lives.