In the practical application of the Bridge Weigh-In-Motion (BWIM) methods, the position of the wheels or axles during the passage of a vehicle is a prerequisite in most cases. To avoid the use of conventional axle detectors and bridge type-specific methods, we propose a novel method for axle detection using accelerometers placed arbitrarily on a bridge. In order to develop a model that is as simple and comprehensible as possible, the axle detection task is implemented as a binary classification problem instead of a regression problem. The model is implemented as a Fully Convolutional Network to process signals in the form of Continuous Wavelet Transforms. This allows passages of any length to be processed in a single step with maximum efficiency while utilising multiple scales in a single evaluation. This allows our method to use acceleration signals from any location on the bridge structure and act as Virtual Axle Detectors (VADs) without being limited to specific structural types of bridges. To test the proposed method, we analysed 3787 train passages recorded on a steel trough railway bridge of a long-distance traffic line. Results of the measurement data show that our model detects 95% of the axles, which means that 128,599 out of 134,800 previously unseen axles were correctly detected. In total, 90% of the axles were detected with a maximum spatial error of 20 cm, at a maximum velocity of $v_{\max }=56.3\mathrm{m}/\mathrm{s}$. The analysis shows that our developed model can use accelerometers as VADs even under real operating conditions.

Force plates are the gold standard for measuring ground reaction forces, but their reliance on equilibrium equations limits them to single-contact measurements. During the double support phase of walking, when both feet touch the ground simultaneously, conventional force plates yield only the combined force and center of pressure. The individual foot contributions cannot be separated. This paper applies influence surface theory to extend force plate analysis beyond equilibrium-based methods. Instead of evaluating global force and moment balances, influence surfaces characterize how loads at different positions contribute to structural responses at individual supports. This enables load identification for multiple simultaneous contacts, with the number of separable loads determined by the available sensor count.The influence surfaces are trained from walking data using Gaussian Process Regression, capturing the real plate behaviour without requiring finite element modelling. To reduce complexity, the inverse problem is addressed sequentially. First, load positions are estimated. Second, the horizontal force components are determined.The method is validated on a 2.5m x 1.0m force plate using static tests with known load positions and walking trials with synchronized instrumented insoles providing the load distribution. The results show that the approach successfully reconstructs positions and three-dimensional ground reaction forces for both contacts, generally achieving accuracy comparable to equilibrium-based single-contact measurements while preserving natural gait.