Roadmap: Now, Next, Later

The robots are coming in stages, and my agents are working every phase at once: research, planning, records. I'm pushing every front, but each phase still has to work before the next one starts.

First big milestone: the 2027 stocker season (Proven method · My own build)

Buy young cattle in spring 2027, graze them through the season with Nofence and the water robot, and in fall sell most of them at a local sale barn. One or two go to a custom-exempt processor for friends and family who bought live shares. Everything below on "Now" and "Next" is in service of this one loop. The full month-by-month plan is on the 2027 milestone page. Related pages: processors near KC and selling at the sale barn.

Now: foundation (Proven)

• Map the land (around 20 acres). Soils, slopes, shade, low spots, where water can come from. NRCS and FSA aerial photos and a soil survey help with paddock layout (MU Extension EQ379).
• Check cell coverage across the pasture. Nofence collars work over AT&T, T-Mobile, and Verizon on LTE-M, and the fence still holds without signal, but alerts and remote moves need some coverage (Nofence coverage guide).
• Secure the perimeter. Nofence itself says most producers keep perimeter and handling fences, and young calves can't wear collars yet (Nofence FAQ).
• Set up a working pressurized water line (hydrant or main), the starting point for portable water (Virginia Tech SPES-710).
• Start the records. One record per animal, EID tags, and dated notes on what I spend and measure.
• Line up a processor early; small processors often book months ahead (UW-Madison Extension). My short list is on processors near KC.
• Stockpile grass and line up hay for winter, and track it (hay network).
• Learn the sale barn. Watch sales at Kingsville and Holton this winter before I buy anything (selling at the sale barn).
• Test the fescue for toxic endophyte. It decides how long the grazing season really is (MU Extension G4669).

Next: the first robots (Early commercial + My own build)

• Nofence collars on the herd. Train on a physical fence first (cattle usually learn the audio cue in 4 to 7 days) and then run virtual paddocks (Nofence).
• Water robot v0 (My own build). A CLEAN-FLOW 350 on a mobile frame, hand-towed first, fed by surface poly pipe with quick couplers. Measure: does it stay clean, does it tip, how fast does it refill.
• Mobile coop v0. A manual eggmobile that follows the cattle a few days behind.
• Baseline soil and forage measurements so I can tell if any of this works.

Later: the robots arrive (My own build)

• Water robot v1. Moves itself to a target spot I set, ideally in step with the Nofence paddock moves.
• Robotic coop. Self-moving coop that follows the herd's path; either one I build myself or a commercial unit I buy (Ukkö ROVA BARN is one real example of the latter).
• One schedule for the whole system. Fence moves, water moves, and coop moves planned together, with an AI assistant handling reminders, records, and the season write-ups.
• Robots that watch gait. A camera on the water robot films cattle walking in to drink, scored against my own locomotion scores (lameness vision).
• Publish the data. Grazing maps, water use, soil results, costs.

Someday: agents organize, robots labor (Speculative)

• Agents as the ranch office. Grazing plans, sale reports, LRP, processor bookings and records. I sign everything (agents and robots).
• Open range with neighbors. Shared grazing across properties, organized by agents, with collars instead of new wire and real fences where the law needs them (open range, agent-organized).
• Direct sales between neighbors, organized by agents (cattle marketplace).
• Seeing the whole industry change (how AI could change the beef industry), while respecting the people already in it (learning the traditions).

What could change the plan

Herd size, budget, cell coverage, winter water (freezing), whether the trough can handle being moved, the fescue test, spring 2027 cattle prices, and processor availability. I'll update this page as those answers come in.