With digital transformation accelerating across the industrial sector, enterprises need to choose the best products to modernize their data infrastructure. This article compares two of those products — TDengine and AVEVA PI System — to help you make informed decisions about your data management strategy.
About AVEVA PI System
PI System, originally developed by OSIsoft and now part of AVEVA, is one of the most widely used industrial data infrastructure platforms for collecting, storing, and organizing time-series data from sensors, equipment, and control systems. It is designed to integrate data from thousands of sources in real time, create contextual models of industrial operations, and deliver this information to operators, engineers, and business systems for monitoring and analysis.
About TDengine
TDengine is an AI-powered data platform designed for industrial applications, combining the high-performance time-series database TDengine TSDB and the AI-native data management platform TDengine IDMP.
With TDengine TSDB handling data ingestion, storage, and processing, and TDengine IDMP providing contextualization, standardization, and AI-powered analytics, TDengine enables industrial enterprises to unlock the true value of their time-series data.
Which Is Best for You?
PI System and TDengine handle many of the same use cases, such as remote operations monitoring, real-time data for support personnel, data science analysis, and data sharing with external partners and applications. However, there are key differences between the products, which can be summarized as follows:
-
Pricing model
-
AVEVA requires customers to purchase its Flex credits in order to use PI System and other services, which can result in dramatically increasing costs for enterprises with tens of thousands or millions of metrics to access.
-
TDengine is designed as an affordable solution for big data, using a SaaS-like pricing model based on compute resources instead of tags, ensuring that customers pay only for what they use. Furthermore, its pricing is transparent, with TDengine Cloud prices even publicly available on the website.
-
-
Cloud support
-
PI System was designed well before the advent of cloud computing, and customers typically need to deploy AVEVA CONNECT data services (formerly Data Hub) in order to make use of the cloud. This means that cloud users have a different experience than on-premises users and makes it more difficult to work with PI System deployments across different sites. In addition, only Windows systems and Microsoft Azure are supported, leaving customers who prefer Linux or AWS behind.
-
TDengine is a cloud-native system and can be easily deployed on Windows or Linux systems at the edge, in public, private, or hybrid clouds, or as a fully managed cloud service in AWS, Azure, or GCP. With its modern design, TDengine is better able to take advantage of the elasticity and flexibility of the cloud, delivering performance and scalability while ensuring an identical experience for cloud and on-premises users.
-
-
AI integration
-
The core PI System does not include native AI or LLM capabilities, and customers that want AI-driven analysis need to integrate with external analytics tools or platforms. This requires development effort on the part of the customer to set up data pipelines, cleansers, and connections, increasing the IT burden on industrial enterprises.
-
TDengine IDMP includes built-in integration with LLMs for zero-query intelligence, delivering AI-generated visualizations and analysis tasks without requiring prompts or queries. Chat BI functionality is also available out of the box for customers who already know what they want to build. Finally, TDengine TDgpt provides AI/ML-based time-series forecasting and anomaly detection within TDengine.
-
-
Data sharing:
-
With PI System and AVEVA CONNECT data services, you can create custom views to define the granularity of shared data, ensuring security and compliance while giving key stakeholders and applications access to needed information. A REST API is offered, but there are few other options.
-
In addition to traditional views, TDengine offers data sharing on a publish-subscribe (pub-sub) model with its data subscription feature, in which applications are notified whenever new data points are inserted into TDengine. Data subscription in TDengine can use a Kafka-like protocol or standard MQTT. This model is better suited for data-intensive AI and other real-time analytics applications than traditional view-based sharing.
-
-
Ecosystem: Although offering many of the same features for many of the same business scenarios, the two products differ on a fundamental level. Due to the closed nature of AVEVA products, PI Systems customers’ choices are more limited in terms of system interoperability and expansion, with only a few vendor-provided offerings. TDengine, built on an open-source core, offers an open ecosystem with standardized interfaces such as JDBC and ODBC, facilitating integration with essentially any third-party product.
-
Developer-friendliness: AVEVA provides a REST API through which applications can communicate with PI System, but other development options are limited. In addition to its own REST API, TDengine delivers a more developer-friendly experience with its client libraries for a wide range of programming languages — Python, R, Java, C#, Go, Rust, and more — including sample code that can be used in custom applications. And because TDengine’s core components are open-source software, developers have an even easier time diving into the code and designing their applications.
Conclusion
PI System has been a key component of industrial data infrastructure for decades, but due to its age may not deliver the same value to industrial enterprises that it did in previous years. Although ripping and replacing PI System deployments may not be advisable, enterprises should begin to consider modern alternatives for new sites.
Enterprises that want to make use of cutting-edge cloud and AI technologies with increased flexibility and reduced operational costs may consider TDengine as a solution for modernizing their industrial data stack. With its open ecosystem, TDengine delivers a wider range of integration and interconnection options than traditional industrial products, never locking you in to a single vendor or holding your data hostage. Efficiency-focused organizations and those interested in the Industrial IoT can benefit greatly from TDengine’s predictable and transparent pricing structure, significantly reducing data infrastructure TCO for deployments with large numbers of devices.
