AI Algorithms to Optimise and Visualise 3D Trilateration Error

Abstract

In this issue we explore a frontier that has now become tangible: collaboration between human and artificial intelligence, leveraged in the development of CAD applications written in AutoLISP, specifically for trilateration-based surveying. The article takes the form of a dialogue with GitHub Copilot, the AI assistant built on advanced artificial intelligence models, which can support developers in the creation of software packages. In this direct exchange, Copilot analyses and describes the trilateration routines it helped generate, revealing the logic underpinning complex algorithms.
In particular, two original tools are presented: the first computes the most probable position of a point in space by comparing and visualising the solutions obtained through different optimisation algorithms; the second transforms the complexity of these calculations into a true three-dimensional “landscape”, where multiple objective functions shape the morphology of the error and make the search for the optimal solution immediate, drawing on a powerful orographic metaphor.
From the final reflections emerge possible strategies for collaboration between algorithms and deterministic, geometry-based approaches, offering both technical and educational insights into the design and understanding of the trilateration process.
This experiment provides a valuable opportunity to observe up close a creative process in which the designer’s intuition merges with the computational power of AI: what follows is the account of this virtual conversation, a snapshot of the future of graphical development.