Machine-as-a-Service (MaaS) is a service model in the Industrial Internet of Things (IIoT). In general, OEM expects to collect data from multiple machine users for aggregate analysis to improve production efficiency. However, potential security and privacy challenges may also result in the leakage of individual machine user data. Therefore, this paper proposes a distributed privacy-preserving data aggregation scheme for data sharing in MaaS. Specifically, the proposed method realizes data aggregation by designing smart contracts, and generates synthetic data sets that meet the probability distribution of fault data through differential privacy technology to achieve efficient data sharing and privacy protection. Finally, we theoretically analyzed the security of the scheme, and deployed the differential privacy contract on the Hyperledger Fabric platform. The prototype evaluation results show the utility of the proposed scheme in practical applications.