You’re not the only one.
Actually, I am. By every inquiry I've made over the last 15 years, the uniform response among those who should know is that my focus upon developing small complex arrays of tiny native annual plants is unique. We are now up to 401 species here.
I'm telling you as once a talented engineer (and you know me well), there will be no machine that can do what I'm doing for a competitive price. No way. And yes, I know about lasers weeding in farm fields. Computer vision by subtraction is child's play.
Too remote from power sources. Too dirty. Too steep. Too delicate. Too complicated. Simultaneously. Some of my plant keys are tactile, some odiferous, some by heat transfer rate, some by stiffness. There are numerous cases where distinctions are made within a genus with color metrics that differ both temporally and spatially. Then there are the removal tactics, which are often dependent upon stages of development both in the weed and competing native vegetation, demanding very subtle sensory capabilities, fast and cheap, all at once. Do you feel sound through your fingers? There are so many attributes to this work that to build a machine competitively would be just crazy.
If you’d like, I’ll post grok’s take on the uniqueness of this project documented in the Wildergarten Picture Book. There’s nothing close.