I have given serious thought to a "robotic lawnmower" similar to the HV Ben linked to. With GPS modules < $50 and many other easy options for relative positioning with very close tolerance, there really isn't any reason a DIYer could not design and build a very reliable robotic lawnmower. Charging it via solar/wind would involve the usual trivial but expensive components. Power consumption could be from 500W to 1.5kW depending on width of cut and travel speed. Assuming "continuous cutting" of a lawn during the course of a week, a 250W PV panel might allow as much as 7 to 21 hours per week of "cutting" during the Summer months.
Charging the unit via the grid would allow considerably larger areas to be maintained. Using 2.8Ah LiOn batteries 20 to a string (74V) would take roughly 2.4 strings per hour of operation @ 500W, or 7 strings per hour of operation @ 1.5kW. For 8 hours of continuous operation 20 strings (400 batteries) @ 500W to 60 strings (1200 batteries) @ 1.5kW. At $7/battery, this represents $2800 to $8400 in batteries alone, but this does not preclude it being a worthwhile investment.
Assuming the mower could maintain ~1 Acre per hour of operation, the device would be capable of mowing > 50 Acres per week. 50 Acres represents roughly 75 homes in a medium size residential neighborhood. If the average homeowner in such a neighborhood spends or pays $25 per week for lawn maintenance this amounts to $1875/week. Assuming lawns need cutting 30 weeks per year, this would be $56,250 per year. A durable, low-maintenance robotic mower could be a very savvy purchase for a planned development, even if the upfront cost was > $75k.
Major golf courses spend millions of dollars annually maintaining grass. Replacing some employees and heavy equipment with small robotic units offers huge cost savings to them.
A robotic lawn mower does not have to carry huge battery banks. In cases where noise is not an issue, a small ICE generator could greatly extend the cutting capacity and reduce the up-front cost.