The second order areas you mention are almost certainly "liquid markets" where software consumers can buy the software production they need at prices they expect the market to require...or to put it another way, the there-is-an-unmet-need-for-software-expertise premise is suspect. Power companies,fleet operators, HVAC controls manufacturers, etc. already have access to people who provide them with software. These things are new to you, but the reason you see them is because they are established.
Solar farms are widely distributed and ordinary across the US, but they tend to be in rural areas off the beaten path...I remember my sense of wonder discovering one unexpectedly in west Georgia (the state) only a couple of years ago. And wind turbines? They are everywhere in the western US, but still amaze me.
I'd put it this way, if you knew a lot of truckers fleet management software would not smell like a greenfield. It's older than GM's onstar service.
When it comes to industrial applications, lots of people have been sitting around thinking what's-next for a long time. They work in organizations capable of funding making-it-happen via ROI.
These aren't bicycle-sized-project. Bicycles can be built by as small team (even one) with jigs and brazing rods and a components closet. A salable electric car can't.
Good luck.