The Industrial Internet-of-Things (IIoT) promises significant improvements for the manufacturing industry by facilitating the integration of manufacturing systems by Digital Twins. However, ecological and economic demands also require a cross-domain linkage of multiple scientific perspectives from material sciences, engineering, operations, business, and ergonomics, as optimization opportunities can be derived from any of these perspectives. To extend the IIoT to a true Internet of Production , two concepts are required: first, a complex, interrelated network of Digital Shadows which combine domain-specific models with data-driven AI methods; and second, the integration of a large number of research labs, engineering, and production sites as a World Wide Lab which offers controlled exchange of selected, innovation-relevant data even across company boundaries. In this article, we define the underlying Computer Science challenges implied by these novel concepts in four layers: Smart human interfaces provide access to information that has been generated by model-integrated AI . Given the large variety of manufacturing data, new data modeling techniques should enable efficient management of Digital Shadows, which is supported by an interconnected infrastructure . Based on a detailed analysis of these challenges, we derive a systematized research roadmap to make the vision of the Internet of Production a reality.

By offering the possibility to already perform processing as packets traverse the network, programmable data planes open up new perspectives for applications suffering from strict latency and high bandwidth requirements. Real-time Computer Vision (CV), with its high data rates and often mission-and safety-critical roles in the control of autonomous vehicles and industrial machinery, could particularly benefit from executing parts of its logic within network elements.In this paper, we thus explore what it takes to bring CV to the network. We present our work-in-progress efforts of implementing a line-following algorithm based on convolution filters on a P4-programmable NIC. We find that by appropriately identifying regions of interest in the image data and by diligently distributing the necessary calculations among the various match/action stages of the ingress-and egress pipelines of the NIC, our prototypical implementation can achieve over 19 decisions per second on 640x480 px grayscale images with filters large enough to guide a small autonomous car through various courses. CCS CONCEPTS• Networks → In-network processing; Middle boxes / network appliances; Programmable networks; • Computing methodologies → Computer vision;

BackgroundWhole genome sequencing has become fast, accurate, and cheap, paving the way towards the large-scale collection and processing of human genome data. Unfortunately, this dawning genome era does not only promise tremendous advances in biomedical research but also causes unprecedented privacy risks for the many. Handling storage and processing of large genome datasets through cloud services greatly aggravates these concerns. Current research efforts thus investigate the use of strong cryptographic methods and protocols to implement privacy-preserving genomic computations.MethodsWe propose Fhe-Bloom and Phe-Bloom, two efficient approaches for genetic disease testing using homomorphically encrypted Bloom filters. Both approaches allow the data owner to securely outsource storage and computation to an untrusted cloud. Fhe-Bloom is fully secure in the semi-honest model while Phe-Bloom slightly relaxes security guarantees in a trade-off for highly improved performance.ResultsWe implement and evaluate both approaches on a large dataset of up to 50 patient genomes each with up to 1000000 variations (single nucleotide polymorphisms). For both implementations, overheads scale linearly in the number of patients and variations, while Phe-Bloom is faster by at least three orders of magnitude. For example, testing disease susceptibility of 50 patients with 100000 variations requires only a total of 308.31 s (σ=8.73 s) with our first approach and a mere 0.07 s (σ=0.00 s) with the second. We additionally discuss security guarantees of both approaches and their limitations as well as possible extensions towards more complex query types, e.g., fuzzy or range queries.ConclusionsBoth approaches handle practical problem sizes efficiently and are easily parallelized to scale with the elastic resources available in the cloud. The fully homomorphic scheme, Fhe-Bloom, realizes a comprehensive outsourcing to the cloud, while the partially homomorphic scheme, Phe-Bloom, trades a slight relaxation of security guarantees against performance improvements by at least three orders of magnitude.

No abstract

No abstract