Electrochemical impedance spectroscopy can reveal aspects of battery behaviour that a single voltage reading cannot, but reliable measurements still depend on making repeatable electrical contact. This paper develops a shared-control robotic workflow for positioning measurement tooling on electric-vehicle battery modules, combining programmed motion with operator-guided final alignment.
What EIS Measures
EIS applies a small alternating electrical perturbation and measures the response at different frequencies. The relationship between voltage and current is a complex impedance: it describes both the relative amplitude and the phase shift. A spectrum therefore contains more information than a single resistance value.
Different processes contribute at different frequencies, which makes EIS useful for investigating cells and their interfaces. Interpretation still depends on test conditions and a suitable model. A spectrum is not, by itself, a remaining-life estimate or an automatic decision that a module is safe to reuse.
Contact is part of this measurement chain. Additional resistance at the probes can change what the instrument observes even when the cell is unchanged. Temperature, state of charge and contact conditions must therefore be controlled or accounted for when comparing measurements.
The Robot and Its Contact Tool
The experimental system uses a KUKA KR20 mounted on a five-metre rail, custom contact tooling and a potentiostat. The rail extends the workspace; the end effector provides the interface between robot positioning and electrical acquisition. Compliant contact elements help accommodate the practical difficulty of aligning the tool with terminals.
Finite element analysis assesses the tooling under modelled loads and constraints. This addresses a mechanical design question, not the accuracy of an entire electrochemical interpretation. The model’s boundary conditions and material assumptions determine what its stress and deformation results can support.
Why the Workflow Uses Shared Control
Preprogrammed approach and retreat motions provide repeatability away from the terminals. Near the contact region, manual admittance control allows the operator’s applied force to guide motion. In simple terms, admittance control turns a measured force into a commanded movement, giving the operator a way to make fine adjustments through the robot.
This is different from a system autonomously detecting every terminal and choosing every contact pose. The reported experiment includes operator-guided alignment and an operator starting acquisition. Describing it as completely autonomous testing would hide the human contribution that makes the demonstrated workflow work.
The paper also describes reduced-speed T1 operation, interlocks and initial mock-up testing. These are aspects of the experimental procedure, not proof that the setup is certified for arbitrary high-voltage packs. Force/torque sensing supports controlled interaction but cannot independently establish electrical safety.
Demonstrated Results and Their Meaning
The work demonstrates EIS on two battery modules connected in series. One reported cell-pair resistance is approximately 1.5 mΩ. That number describes the measured case: it is not a general boundary separating healthy and degraded batteries.
The central result is the integration of robotic positioning and measurement in a laboratory workflow. It shows how contact tooling, robot motion and operator guidance can be combined so that electrical characterisation becomes part of a robotic process.
A stronger production claim would need additional evidence on repeated contacts, measurement repeatability, cycle time, tool wear and different terminal geometries. Quantified reductions in operator exposure also require an appropriate comparison rather than being inferred simply because a robot is present.
From a Demonstration to a Reproducible Assessment
A meaningful reproduction would keep the mechanical and electrical conditions aligned: terminal geometry, contact force, tooling, robot/controller configuration, instrument calibration, state of charge and temperature all influence the outcome. Comparing spectra under matched conditions would help distinguish variations in contact from variations in the battery.
The interesting research question is not merely whether the robot reaches a position. It is whether the complete workflow produces repeatable, interpretable measurements without transferring positioning errors into the electrical result. Physical work belongs in a suitably equipped laboratory with qualified supervision and its established battery-safety procedures.
Testing and Real-World Use
This could be tested in a qualified laboratory by repeating the robot-guided contact procedure under matched battery conditions and comparing the resulting spectra with a controlled reference measurement. In battery characterisation and remanufacturing research, the workflow could reduce repetitive probe positioning while retaining human judgement for uncertain alignment and measurement conditions.
Paper and Further Figures
The full open-access article includes the contact-tool CAD, control workflow, electrical measurements and tooling analysis. These figures document the experiment; they should not be interpreted as a universal procedure for working on live battery packs.