Interactive Visual Analysis of Robot Base Placement and Trajectory Optimization

Abstract

Optimization procedures for industrial robot base placement and trajectory planning typically result in the selection of a single solution and provide limited support for understanding the broader solution space or reasoning about trade-offs between conflicting objectives. The high-dimensional data, including spatial dependencies and derived performance metrics, requires tailored visual analysis support. To address this gap, we present ROSS, a design study for exploring ensembles of optimization results for robot base placement and trajectory planning. We propose a spatially-linked parallel coordinates plot to increase the analytical prominence of x/y base positions within the high-dimensional data. We gathered formative feedback from industry practitioners and conducted six summative semi-structured interviews. Preliminary findings indicate that complementary visual encodings support distinct analysis strategies and reasoning about trade-offs.