Edge-computing enhanced privacy protection for industrial ecosystems in the context of SMEs

Abstract

The ongoing transformation of the manufacturing landscape introduces new business opportunities for enterprises but also brings new challenges with it. Especially small-and medium-sized companies (SMEs) require an increasing effort to stay competitive. Data produced on the shop-floor can be harnessed to conduct analyses useful to plant operators, e.g., for optimization of production capabilities or for increasing plant security. Therefore, we propose a privacy-preserving edgecomputing architecture to facilitate a platform for utilizing such applications. Our approach is motivated by requirements from SMEs in Germany, e.g., protection of intellectual property, and employs suitable privacy models. We demonstrate the viability of the proposed framework by evaluation of uses cases for machine chatter optimization and anomaly detection within plants.

Leveraging Edge Computing and Differential Privacy to Securely Enable Industrial Cloud Collaboration Along the Value Chain

“…In [1], an edge computing based scheme for industrial ecosystems was presented with an included privacy model. An evaluation was performed on the given use case of the chatter stability analysis as well, whereas the effects of the privacy model on the machine tool data are not discussed.…”

“…As a privacy model, DP was chosen. The usage of DP is motivated by an initial requirements analysis [1]. For the presented application, DP offers the advantage that it can be used locally by a data producer, allowing full control over the data before leaving the company's premises.…”

“…Although machine tool operators benefit from these optimizations, they hesitate to share their data with third parties due to concerns regarding the loss of intellectual property [1,2]. This dilemma is underpinned by the increasing emergence of third-party risks, such as attacks on the supply and value chain respectively [3,4].…”