Controlling from cDAQ to GPIB and PXI with Natural Language Commands

Supporting Developers from Equipment Selection to Acceptance Code Analysis

NI Nigel™ AI guides engineers through necessary equipment and development procedures when measurement objectives are explained in natural language, and even generates LabVIEW code to control actual measurement instruments. Unlike general question-and-answer tools, NI Nigel™ AI is specialized in test system design, measurement methodology, hardware configuration, and data analysis—not a tool replacing engineers, but a new colleague amplifying test expertise.

NUBICOM, an official distributor of NI, held the 'NI Nigel™ AI Launch Event' on the 13th at the Ace High Tech City Convention Hall in Yeongdeungpo, Seoul, and unveiled the key features and use cases of Nigel AI.

Developers and engineers attending the event confirmed the process of applying AI to test and measurement tasks through introduction sessions, live demos, and hands-on experience.

Manager Kim Ki-ryung of NUBICOM, who presented at the event, explained that 'NI Nigel™ AI' is an AI developed based on NI's test and measurement expertise, answering questions about test system design, measurement methodology, and data analysis, and understanding the context of projects under development within the LabVIEW environment to suggest necessary functions and programming directions.

Particularly noteworthy is that it provides answers more suited to test applications and NI hardware compared to general-purpose AI.

The most highlighted feature is automatic LabVIEW VI generation.

When engineers input the measurement target, sensors, equipment, and desired results in natural language, Nigel AI analyzes the requirements and creates block diagram code.

Connected equipment recognition and measurement instrument control code generation are also key features.

Nigel AI confirms the list of NI hardware connected to the PC and third-party measurement instruments with installed drivers, supporting program development that reflects equipment characteristics.

Nigel AI is also utilized in existing code analysis and development guidance. When a LabVIEW block diagram is attached, it analyzes and explains the controls and indicators used, program structure, and behavior by state.

When taking over code written by other developers or improving existing test programs, time spent understanding code structure can be reduced. Required functions, design patterns, and improvement directions can also be inquired through conversation.

Test system configuration consulting is also a core feature of Nigel AI. When the measurement target, signal type, and requirements are input, it provides step-by-step guidance on necessary hardware, drivers, toolkits, key functions, and screen layout components.

In the live demo, a program was created to simultaneously measure temperature and vibration using cDAQ.

When Kim Jin-ho, a manager at NUBICOM, input the measurement objective and equipment conditions as a prompt, Nigel AI generated a VI that configured temperature and vibration input as a single task.

Instead of processing the two modules in separate loops, it was configured to bundle them into a single task to read synchronized data.

When the temperature sensor was held in the hand, the measurement value changed, and when the fan operated, the vibration graph responded.

Subsequently, Nigel AI created a VI controlling a power supply and multimeter connected via GPIB.

When the program commanded 5V output, the power supply operated, and the multimeter transmitted measurement data to the PC. In the PXI system, it confirmed the mounted signal generator and oscilloscope modules and generated a VI using both instruments.

Manager Kim Jin-ho explained that the more specific the desired measurement conditions, front panel configuration, and equipment operation method are input, the higher quality results can be obtained.

The assistance provided by Nigel AI is not limited to automatic code generation. First, developers designing new test systems can receive preliminary configuration consulting.

For example, when measuring motor temperature and vibration, inputting the measurement target, available equipment, required results, and considerations provides guidance on necessary NI hardware drivers and toolkits, locations of key functions, and controls and indicators to place on the front panel.

When unsure where to start, Nigel AI can be asked to first organize the considerations.

Help can also be obtained when handling measurement instruments for the first time.

Nigel AI confirms the list of equipment connected to the PC and provides guidance on what specifications and purposes the equipment is used for.

When it is difficult to know what measurement instruments are connected to old equipment, or when it is necessary to verify dedicated drivers and function usage rules for each PXI module, it can be used as a starting point for equipment exploration and learning.

It also supports understanding the structure of code handed over from other developers.

At the event, when a continuous measurement LabVIEW template was attached to Nigel AI, it distinguished and explained the start, stop, and configuration controls, data indicators, overall design structure, and behavior by state.

Previously, developers had to manually trace and analyze the block diagram, but with Nigel AI, developers can first receive a summary of how the code operates and then focus on reviewing necessary parts.

It can also be used for draft development and learning.

Nigel AI proposes design patterns suited to the purpose, tells the required functions and their locations, and writes step-by-step development guides or process charts.

Developers with limited LabVIEW experience can secure the skeleton of a measurement program instead of starting with a blank block diagram, and experienced developers can request revisions based on NI example programs to reduce repetitive initial implementation work.

According to internal experience shared by NI, code generation using Nigel AI confirmed time savings of 5 to at least 10 times or more compared to traditional engineer development. However, currently the focus is on generating new single measurement VIs, with limitations in automatic merging of multiple VIs, object-oriented structures, sophisticated UI layout, and FPGA and real-time code generation. The generated results also require developer verification and supplementation.

Nevertheless, there are clear reasons to use Nigel AI.

It continuously supports equipment exploration, system configuration, development procedure establishment, draft code generation, and existing program analysis within LabVIEW.

It reduces time spent on repetitive implementation and searching, helping developers focus on core tasks of determining the validity of measurement conditions and the meaning of data.

NI plans to further develop Nigel AI in the future to connect not only LabVIEW but also TestStand, DIAdem, VeriStand, and other test software platforms as an agent.