Search

This article presents a circuit model that is able to capture the full nonlinear behavior of an asymmetric electrostatic transducer whose dynamics are governed by a single degree of freedom. Effects such as stress-stiffening and pull-in are accounted for. The simulation of a displacement-dependent capacitor and a nonlinear spring is accomplished with arbitrary behavioral sources, which are a standard component of circuit simulators. As an application example, the parameters of the model were fitted to emulate the behavior of an electrostatic MEMS loudspeaker whose finite-element (FEM) simulations and acoustic characterisation where already reported in the literature. The obtained waveforms show good agreement with the amplitude and distortion that was reported both in the transient FEM simulations and in the experimental measurements. This model is also used to predict the performance of this device as a microphone, coupling it to a two-stage charge amplifier. Additional complex behaviors can be introduced to this network model if it is required.

With the increase in applications of machine learning methods in drug design and related fields, the challenge of designing sound test sets becomes more and more prominent. The goal of this challenge is to have a realistic split of chemical structures (compounds) between training, validation and test set such that the performance on the test set is meaningful to infer the performance in a prospective application. This challenge is by its own very interesting and relevant, but is even more complex in a federated machine learning approach where multiple partners jointly train a model under privacy-preserving conditions where chemical structures must not be shared between the different participating parties. In this work we discuss three methods which provide a splitting of a data set and are applicable in a federated privacy-preserving setting, namely: a. locality-sensitive hashing (LSH), b. sphere exclusion clustering, c. scaffold-based binning (scaffold network). For evaluation of these splitting methods we consider the following quality criteria (compared to random splitting): bias in prediction performance, classification label and data imbalance, similarity distance between the test and training set compounds. The main findings of the paper are a. both sphere exclusion clustering and scaffold-based binning result in high quality splitting of the data sets, b. in terms of compute costs sphere exclusion clustering is very expensive in the case of federated privacy-preserving setting. Supplementary Information The online version contains supplementary material available at 10.1186/s13321-021-00576-2.

Electrostatic actuators are of particular interest for microsystems (MEMS), and in particular for MEMS audio transducers for use in advanced true wireless applications. They are attractive because of their typically low electrical capacitance and because they can be fabricated from materials that are compatible with standard complementary metal-oxide semiconductor (CMOS) technology. For high audio performance and in particular low harmonic distortion (THD) the implementation of the push-pull principle provides strong benefits. With an arrangement of three electrodes in a conjunct moving configuration on a beam, we demonstrate here for the first time a balanced bending actuator incarnating the push-pull principle operating at low voltages. Our first design already exhibits a harmonic distortion as low as 1.2% at 79 dB using a signal voltage of only 6 Vp and a constant voltage of only ±10 Vdc in a standard acoustic measurement setup. Thus, exceeding our previously reported approach in all three key performance indications at the same time. We expect that our novel electrode configurations will stimulate innovative electrostatic actuator developments for a broad range of applications. In this paper we report the basic theory, the fabrication and the performance of our novel actuator design acting as an audio transducer.

To apply federated learning to drug discovery we developed a novel platform in the context of European Innovative Medicines Initiative (IMI) project MELLODDY (grant n{\deg}831472), which was comprised of 10 pharmaceutical companies, academic research labs, large industrial companies and startups. The MELLODDY platform was the first industry-scale platform to enable the creation of a global federated model for drug discovery without sharing the confidential data sets of the individual partners. The federated model was trained on the platform by aggregating the gradients of all contributing partners in a cryptographic, secure way following each training iteration. The platform was deployed on an Amazon Web Services (AWS) multi-account architecture running Kubernetes clusters in private subnets. Organisationally, the roles of the different partners were codified as different rights and permissions on the platform and administrated in a decentralized way. The MELLODDY platform generated new scientific discoveries which are described in a companion paper., Comment: 9 pages, 4 figures, to appear in AAAI-23 ([IAAI-23 track] Deployed Highly Innovative Applications of AI)

Federated multi-partner machine learning can be an appealing and efficient method to increase the effective training data volume and thereby the predictivity of models, particularly when the generation of training data is resource intensive. In the landmark MELLODDY project, each of ten pharmaceutical companies realized aggregated improvements on its own classification and/or regression models through federated learning. To this end, they leveraged a novel implementation extending multi-task learning across partners, on a platform audited for privacy and security. The experiments involved an unprecedented cross-pharma dataset of 2.6+ billion confidential experimental activity data points, documenting 21+ million physical small molecules and 40+ thousand assays in on-target and secondary pharmacodynamics and pharmacokinetics. Appropriate complementary metrics were developed to evaluate predictive performance in the federated setting. In addition to predictive performance increases in labeled space, the results point towards an extended applicability domain in federated learning. Increases in collective training data volume, including by means of auxiliary data resulting from single concentration high-throughput and imaging assays, continued to boost predictive performances, albeit with saturating return. Markedly higher improvements were observed for pharmacokinetics and safety panel assay-based task subsets.