Product Overview

The Siemens 6ES7556-1AA00-0AB0 SIMATIC S7-1500 Neural Processing Unit (TM NPU) is an artificial intelligence technology module used to execute neural networks and process sensor data within industrial automation systems. Unlike conventional S7-1500 I/O modules, which primarily acquire signals or control field devices, the TM NPU handles computational tasks involving image data, neural network inference, and application-specific data processing.

The module is particularly useful in machine vision applications where conventional PLC logic alone is not sufficient for image-based inspection or classification. It can process data from connected cameras as well as information supplied by the S7-1500 CPU user program. The results can then be exchanged with the controller through the process image.

A typical application is an automated quality inspection station on a production line. A camera captures images of manufactured components, and a trained neural network evaluates the image to identify defects, classify products, or determine whether a component meets specified visual criteria. The TM NPU executes the neural network and makes the resulting data available to the PLC, which can use the evaluation results to control sorting equipment or trigger an inspection response.

The module uses an Intel Movidius Myriad X processing unit with two LEON processors operating at 700 MHz. This hardware architecture is intended for neural network processing rather than general-purpose PLC control. It allows image-related computational tasks to be handled separately from the main S7-1500 CPU, leaving the controller responsible for machine sequencing, interlocking, and actuator control.

The TM NPU provides one Gigabit Ethernet interface and one USB 3.1 interface. Cameras can be connected through the Ethernet or USB interface, depending on the supported camera type and application configuration. The Ethernet interface supports GigE Vision, making it suitable for industrial camera applications. The USB interface is also available for camera testing and commissioning, although it is not intended for productive camera operation.

Another useful feature is its integrated MicroPython interpreter, which provides an open software environment for application-specific processing. Engineers can load application files and neural network models onto the module and develop processing workflows according to the requirements of the inspection or classification task.

Application files and neural network data are transferred using a compatible SIMATIC Memory Card. This allows the module to load the software and model files required for a particular machine vision application.

For example, in an electronic component manufacturing line, the camera can capture images of circuit board assemblies while the TM NPU evaluates component placement or identifies visual irregularities. The S7-1500 CPU receives the processed information and determines whether the product should continue through the production process or be diverted for further inspection.

The 6ES7556-1AA00-0AB0 is therefore best suited to automation systems that need neural network processing alongside conventional PLC control, particularly in machine vision, automated inspection, product classification, and sensor-data evaluation.

Technical Specifications

Product Features

• Neural Network Processing: Loads and executes trained neural networks for industrial image analysis and classification.
• Intel Myriad X Processor: Uses a dedicated neural processing architecture with two LEON processors operating at 700 MHz.
• MicroPython Environment: Provides an open software architecture for application-specific processing.
• Gigabit Ethernet Interface: Supports industrial camera communication through a 1000 Mbps RJ45 connection.
• USB 3.1 Connectivity: Provides a USB Type-A interface for compatible devices and commissioning activities.
• GigE Vision Support: Allows connection of compatible industrial vision cameras.
• CPU Data Exchange: Exchanges application data with the S7-1500 CPU through the process image.
• SIMATIC Memory Card Support: Loads application-specific software and neural network files from a compatible memory card.
• Separate Processing Architecture: Handles AI-related processing independently of the main PLC control program.
• Camera Data Evaluation: Processes image data supplied by supported camera configurations.
• Integrated Diagnostics: Provides RUN, ERROR, MAINT, LINK, and power supply status indicators.
• Industrial Mounting: Installs on the SIMATIC S7-1500 mounting rail and supports horizontal or vertical mounting.

Working Principle

The 6ES7556-1AA00-0AB0 operates as a dedicated neural network processing unit within the automation system. It receives image or application data, executes the configured neural network, and exchanges the processing results with the S7-1500 CPU.

The basic data flow is:

Industrial Camera / Sensor Data → TM NPU → Neural Network Processing → Processed Results → S7-1500 CPU → Machine Control

Unlike a conventional digital input module, the TM NPU does not simply report whether a signal is high or low. Its primary task is to process more complex data, particularly image information, and extract results that can be used by the machine control program.

Camera Data Acquisition

The process begins when a compatible camera captures an image of the inspected product or machine component. The image data is transferred to the TM NPU through its Ethernet or USB interface, depending on the camera and application configuration.

For industrial production, Gigabit Ethernet cameras using GigE Vision are particularly relevant. The USB interface can also be used for supported camera configurations during testing and commissioning.

Neural Network Execution

Once the image or application data is available, the TM NPU executes the neural network loaded into the module.

For example, in a visual inspection application, a trained model can evaluate an image to identify surface defects, recognize component types, or classify products according to predefined characteristics.

The Myriad X processing unit handles the neural network workload, separating this processing activity from the main PLC control tasks.

Application-Specific Processing

The integrated MicroPython interpreter provides an environment for application-specific processing. Engineers can load the required application files and neural network models onto the module through a compatible SIMATIC Memory Card.

This allows the processing workflow to be adapted to the inspection task. Different applications may use different neural network models depending on the products, camera arrangement, and inspection criteria.

Data Exchange with the S7-1500 CPU

After processing, the relevant application data is exchanged with the S7-1500 CPU through the process image.

For example, a visual inspection application may generate classification information indicating whether a product passes inspection or requires rejection. The CPU receives the application data and uses it within the machine control program.

The controller can then activate a sorting mechanism, stop the production process, or record the inspection result according to the configured logic.

Integration with Industrial Vision Systems

The TM NPU can also process data from supported industrial cameras through its Gigabit Ethernet interface. This is useful in applications where image acquisition and neural network evaluation form part of a continuous production process.

The USB camera connection has a different role: it is intended for testing and commissioning rather than normal productive operation. Engineers should therefore select a supported GigE Vision camera for production applications where continuous camera operation is required.

By separating image processing from machine control, the TM NPU allows the S7-1500 CPU to concentrate on sequencing, communication, and equipment coordination while the technology module handles neural network workloads.

Role in the SIMATIC S7-1500 Control Architecture

The 6ES7556-1AA00-0AB0 occupies a specialized processing position within the SIMATIC S7-1500 architecture. It extends the automation system with neural network execution and application-specific data processing without replacing the main PLC CPU.

A typical system arrangement follows this structure:

Industrial Camera → TM NPU → Neural Network Evaluation → S7-1500 Process Image → CPU → Machine Control / Sorting Equipment

In this arrangement, the TM NPU handles computational tasks that are different from the CPU’s normal PLC responsibilities.

For example, a production line may use the NPU to classify components based on their visual appearance. The CPU receives the classification results and determines the next machine action. This separation allows image analysis and machine sequencing to operate as distinct but coordinated tasks.

The module can also be integrated into an ET 200MP distributed configuration using compatible interface modules. This provides additional installation possibilities when the processing unit needs to be located away from the central S7-1500 rack.

The TM NPU therefore acts as an AI processing extension to the automation system rather than a replacement for conventional PLC processing or standard I/O modules.

Industrial Applications

Installation and Maintenance

1. Verify the Module Type: Confirm that Siemens 6ES7556-1AA00-0AB0 is the required TM NPU and that the intended application requires neural network processing.
2. Check CPU Compatibility: Verify the S7-1500 CPU version and the supported central or distributed installation configuration.
3. Check Engineering Software: Confirm that the installed STEP 7 / TIA Portal version and required Hardware Support Package support the module firmware.
4. Inspect the Module: Examine the housing, mounting contacts, Ethernet connector, USB port, and memory card slot for damage or contamination.
5. Prepare the Mounting Rail: Ensure that the S7-1500 mounting rail is correctly installed and that sufficient cabinet space is available.
6. Install the Module: Mount the TM NPU in the designated position and verify that the module is properly seated.
7. Connect the Power Supply: Connect the 24 V DC supply and verify that the voltage remains within the permitted operating range.
8. Connect the Camera Interface: Connect the supported industrial camera through the appropriate Ethernet or USB interface. Use a compatible GigE Vision camera for productive Ethernet-based operation.
9. Install the SIMATIC Memory Card: Insert the specified compatible memory card containing the required application files and neural network data.
10. Configure the Module: Use STEP 7 / TIA Portal to configure the input and output address areas, Ethernet interface, USB interface, and other required parameters.
11. Test Neural Network Processing: Verify camera communication, application loading, neural network execution, and data exchange with the S7-1500 CPU.
12. Monitor Diagnostics: Check the RUN, ERROR, MAINT, LINK, and PWR indicators during commissioning and normal operation. Record any diagnostic messages and configuration changes.

Compatible System Components

Recommended Related Models

Key Advantages

• Executes neural networks directly within the S7-1500 automation architecture.
• Uses the Intel Myriad X processing unit for AI-related workloads.
• Supports image processing and application-specific data evaluation.
• Provides Gigabit Ethernet connectivity for compatible industrial cameras.
• Includes USB 3.1 connectivity for supported devices and commissioning.
• Supports an integrated MicroPython programming environment.
• Loads application files and neural network models through a SIMATIC Memory Card.
• Exchanges processed data with the S7-1500 CPU through the process image.
• Separates AI processing tasks from conventional PLC control operations.
• Supports compatible central and distributed automation configurations.

Technical FAQs

What is the main function of the Siemens 6ES7556-1AA00-0AB0 TM NPU?

The TM NPU is used to execute neural networks and process application data within SIMATIC S7-1500 automation systems. Typical tasks include image classification, visual inspection, and sensor-data evaluation. The processed results can be transferred to the S7-1500 CPU for further machine control decisions.

Which cameras can be connected to the TM NPU?

The module provides one Gigabit Ethernet interface and one USB 3.1 interface. Compatible GigE Vision cameras can be connected through Ethernet for industrial operation. The USB interface supports camera testing and commissioning, but USB camera operation is not intended for productive use. Camera compatibility and supported image formats should be checked before configuring the application.

How are neural network models loaded into the 6ES7556-1AA00-0AB0?

Application-specific files, including neural network models, are loaded using a compatible SIMATIC Memory Card. The specified memory card family is 6ES7954-8LPxx-0AA0 with a capacity of 2 GB. The application files are prepared for the intended processing task and made available to the module through the memory card.

Can the TM NPU replace the S7-1500 CPU for machine control?

No. The TM NPU is an additional processing module, not a replacement for the PLC CPU. It handles neural network execution and application-specific processing, while the S7-1500 CPU continues to execute machine sequences, interlocks, communication, and actuator control. The two devices exchange application data through the process image.