Software Defined Vehicles
Interview with Thiruvenkata Babuji Baskaran, Molex
“A digital twin can predict current-related failure modes very well”
As rising system complexity challenges validation in automotive development, digital twins are becoming a key enabler. Thiruvenkata Babuji Baskaran, Molex, explains how virtual validation transforms connector design and testing.
As software-defined vehicles increase system complexity, traditional validation approaches are reaching their limits, particularly for components such as electrical connectors that must meet strict reliability and durability requirements. Physical testing alone is no longer sufficient to explore the growing number of design permutations and operating conditions.
Thiruvenkata Babuji Baskaran, Predictive Functional Leader at Molex, focuses on advancing high-fidelity digital twin technologies for connector development. With extensive experience in simulation, material characterisation and structural analysis, he works on enabling virtual validation strategies that reduce development time while maintaining engineering confidence.
At the Bordnetzkongress 2026 in Ludwigsburg, Baskaran presents how digital twins can be used to predict failure modes, improve design robustness and potentially reduce the need for physical validation phases. In the following interview, he explains how simulation, data and engineering workflows must evolve to make this approach scalable.
ADT: Looking ahead five years, what will be the single biggest challenge for the wiring harness and EDS industry and why?
Advertisement
Baskaran: Managing increasing electrical power density and system complexity under aggressive cost and timing pressure. Rapid electrification, from 48 V to high-voltage architectures, higher current density in compact packaging, weight reduction versus durability trade-offs, and shorter OEM development cycles all contribute to this challenge.
What is the one capability companies must build now to make data-driven harness development scalable?
The real enabler is not more simulation, but a closed-loop data ecosystem in which every simulation and test becomes reusable engineering intelligence, supported by a simulation process data management system. Such an infrastructure enables scalable digital engineering workflows through automated cross-functional integration, as well as traceable and reusable engineering data, which becomes the foundation for AI-driven solutions. It also supports AI-assisted design exploration to evaluate large design spaces and accelerate decision-making.
Which connector failure modes can a digital twin realistically predict today and which remain largely empirical?
A digital twin can predict connector mechanical behaviour and current-related failure modes very well. However, several aspects remain largely empirical, including fretting corrosion, micro-motion-induced contact resistance drift, and long-term plating degradation.
What inputs are essential for credible virtual validation, such as materials data, contact models, manufacturing tolerances or load cases?
Credible virtual validation depends on a comprehensive set of inputs. Materials data must include temperature-dependent modulus and yield, creep and stress relaxation curves, ageing degradation factors and anisotropic properties, for example in glass-fibre-reinforced plastics. Contact modelling requires assumptions about surface roughness, appropriate friction model selection, non-linear contact stiffness and friction coefficients that accurately reflect the system. Manufacturing effects must also be considered, including injection moulding simulation, fibre orientation, weld line locations, dimensional tolerance stack-up and residual stress from forming. In addition, realistic load cases such as thermal cycling ranges, mating and unmating cycles, and assembly conditions are essential.
How do you correlate simulation with test results and what is your approach to model calibration and confidence?
Advertisement
A structured approach is required to ensure correlation and model confidence. This typically begins with baseline geometry validation, followed by calibration of the material model using coupon tests. Sub-assemblies are then validated before moving to system-level validation. Throughout this process, force prediction error is tracked, sensitivity studies are used to understand dominant variables, and confidence bands are defined. Confidence in the model is further supported by gauge repeatability and reproducibility, clearly defined assumptions and quantified error margins.
Where does digital twin adoption create the biggest time saving, such as in the concept phase, design iterations or qualification planning?
While the qualification phase sees efficiency gains, the concept phase represents a more fundamental transformation. Time savings include a 30 to 50 per cent reduction in late design changes, faster convergence during the design verification phase and fewer redesign loops during production validation.
Finally, what do you personally hope to take away from the Bordnetzkongress 2026 in Ludwigsburg?
The main objective is to gain a clearer understanding of evolving OEM and Tier supplier expectations. Simulation and digital engineering address part of the challenge, but the key lies in aligning these capabilities with real industry needs, whether in faster development cycles, higher power architectures or improved reliability. Conferences such as the Bordnetzkongress help clarify where these priorities are heading and how capabilities must evolve to support them.