[Evidlo]

I published a paper in this area a few years ago (but with a different lens design), so I think I can answer.

In conventional diffractive optics, the focal plane (where the sensor should be placed for a focused image) is wavelength-dependent. This means that you receive an in-focus image of the scene for a specific wavelength λ_i when your sensor is located a distance d_i from the lens [0]. By varying d_i you can scan the scene for all wavelengths even if your sensor is monochromatic. You get blurry versions of all other wavelengths as well, but the idea is to use computational techniques to separate these out.

Here, they've designed an optic which tries to eliminate the wavelength dependence of the lens focal plane. Instead of you place an RGB sensor at this one focal plane to get all the wavelengths.

0: https://ieeexplore.ieee.org/abstract/document/9191355?figure...

[jcims]

Could these help with field curvature as well? To make the focal plane more, uh, planar?