Antenna design requires significant human domain expertise and manual iterative design.
Engineering design · United States
Executive brief
The operating-model shift, in one view.
Shifting humans from designing solutions to designing constraints enables computational optimization to exceed human benchmarks.
AI value · Design turnaround time (re-design)
Verified
Less than one month
No additional attribution caveat published.
Before
Manually design antenna (e.g., quadrifilar helix) based on mission requirements → Fabricate and test antenna in anechoic chamber
After
Define fitness function and mission constraints → Search design space and automatically produce novel antenna designs → Fabricate and physically test the evolved antenna for compliance
Human boundary
Algorithm generates the design; human engineers retain authority to confirm compliance via physical testing.
Why it matters
Antenna design requires significant human domain expertise and manual iterative design.
How the work changed
Before
How the work ran before the change.
Step 1 of 2
Antenna Designer
Manually design antenna (e.g., quadrifilar helix) based on mission requirements
Step 2 of 2
Test Engineer
Fabricate and test antenna in anechoic chamber
ControlPhysical testing
What changed
Antenna design requires significant human domain expertise and manual iterative design.
Decision rightHuman moves from creator to judge
After
How the same work runs now.
Step 1 of 3
Human Engineer
Define fitness function and mission constraints
Step 2 of 3
Evolutionary Algorithm
Search design space and automatically produce novel antenna designs
Step 3 of 3
Test Engineer
Fabricate and physically test the evolved antenna for compliance
ControlAnechoic chamber testing for final approval
Process model built from the published workflow evidence for NASA. Every step, actor, and control appears in full below.Every step, actor, and control
Exception path
If requirements change, engineers adjust the fitness function and re-evolve the design.
Work removed
Manual iterative calculation of electromagnetic interactions
Decision authority
Algorithm generates the design; human engineers retain authority to confirm compliance via physical testing.
Before
01
Antenna Designer
Manually design antenna (e.g., quadrifilar helix) based on mission requirements
Control: Not published.
02
Test Engineer
Fabricate and test antenna in anechoic chamber
Control: Physical testing
After
01
Human Engineer
Define fitness function and mission constraints
Control: Not published.
02
Evolutionary Algorithm
Search design space and automatically produce novel antenna designs
Control: Not published.
03
Test Engineer
Fabricate and physically test the evolved antenna for compliance
Control: Anechoic chamber testing for final approval
Work that left the path
Manual iterative calculation of electromagnetic interactions
Human role before
Manually conceptualize and optimize antenna geometry.
Human role after
Set requirements (fitness function) and test the AI-generated physical design.
AI roleIteratively evolve and optimize antenna geometry to meet the fitness function.
Outcomes
Design turnaround time (re-design)
Verified
Time and labor intensive→Less than one month
ST5 mission adjustment phase · One antenna re-design
No additional attribution caveat published.
Antenna efficiency (dual mounted)
Verified
38%→93%
Pre-launch testing · Two antennas mounted on spacecraft
Compared two hand-designed QHAs against two evolved antennas.
What leaders can reuse
Anti-pattern
Assuming evolutionary computation is out of scope for AI cataloging (Editorial flag raised).
Questions
01Where in our R&D could we specify requirements and let AI search the design space?
Portability conditions
Clear fitness function can be defined
Ability to physically test the result
Reputation risk
low
Evidence and authority
What the public record supports.
Current · updated
1 primary; publication outcomes are verified.
Bundle 1.0.0 · reviewed 2026-08-23 · stable ID 000d0bb757d8a7ce
