AI development is constantly morphing, changing, and evolving with new developments emerging in a wide range of areas with breathtaking speed. This is true for companies ranging from small startups to large, established ones like Siemens. That said, even in a space of constant change there are some key areas that stand out for the value they provide. At Siemens Digital Industries AI development focus could be categorized into a few broad focus areas:
- smart user interfaces
- design space exploration
- the AI-enabled shop floor, and
- specific vertical use cases.
Below some examples of these key use cases are detailed.
In the area of smart interfaces, products and services can be enhanced by AI for the benefits of productivity gains that come from “easier use.” AI command prediction systems help direct a design or simulation engineer by placing the relevant commands at their fingertips; some examples from Siemens Xclerator include Adaptive UI in NX, AI with Teamcenter Assistant and Mendix AI assisted low-code programming.
Related to this area are conversational software and chatbots. These are digital companions that can revolutionize the access to information for everyone from plant owners and technicians, to designers and QA testers.
Computer aided engineering (CAE), simulation and testing software are all complex applications with steep learning curves. Even for experts, powerful software can be difficult to navigate because common functionality is layered in complex menu trees. The Siemens NX team addressed this issue with their AI-enabled smart UI and command predictions. By training an AI with data gathered from the way a person interacts with the software, the software can present a selection of functions needed at any given time. This technique is further extended into full command prediction, where the software can recommend actions, methods and best practices based on completed projects and input. Integrating this type of functionality can drastically flatten the learning curve while capturing valuable design knowledge.
Design space exploration
Design space exploration is the art of using simulation to iterate through many design variants to find and chose the ones that best satisfy design constraints. This applies to product designs of course, but the same applies to the design of a plant, a production line or a CNC (computer numerical control) machine in the factory. Many of the steps in the design space exploration process can be slow or cumbersome for a human to complete, making AI an optimal partner for accelerating the process.
High-fidelity simulations are an important part of the design process, and they are notoriously slow. By leveraging AI and machine learning, Siemens Software engineers are developing methods to speed up simulation runs. For example, the Simcenter suite of tools uses AI/ML-driven, reduced-order models and surrogate models to carefully select a sampling of simulation data, after which they can be used to infer accurate simulation results in real time across the entire domain of interest. These techniques reduce the total number of simulations needed during the design process.
Thanks to the highly parallel nature of microprocessors and the vast quantities of highly accurate data available from SPICE simulations, chip design and validation were prime candidates to reap the early benefits of machine learning. The Siemens EDA tool Solido has been using machine learning to reduce the number of simulations required to validate a chip design from billions down to just a few thousand.
AI today also supports smart sampling and testing. Sampling is the art of selecting test cases and related data so that significant coverage of the areas to test is reached while choosing the smallest number of tests possible to reach the goal. Applied to the problem of design space exploration, this can help dramatically accelerate the process.
Vertical AI solutions for solving industry-specific challenges
Siemens had the opportunity to develop several specific AI solutions in-house such as closed-loop analytics for in-line quality prediction of printed circuit boards, an AI solution that has been developed and hardened in our own electronics factories. Further AI models help with controlling the die-casting process and the silicon-ingots grow process for the semiconductor and solar industries.
Another example is Supplyframe, a platform designed to provide supply-chain transparency in the electronics part industry from both sides: the collection of input data from thousands of different companies and the analytics to help customers with sourcing strategy recommendations. These capabilities would not be possible without the heavy use of AI.
AI offers a breadth of solutions
AI isn’t an industry. It is a digital assistant, the missing piece to realize a new product, a tool to automate a boring task, or the key piece to running an autonomous smart factory – one powerful piece of a puzzle that spans across every industry. AI is finding roles big and small across every segment of daily life and industry as experts continue to shape and mold it into a powerful tool that can revolutionize the way we interact with technology and the world around us.
Siemens Digital Industries Software helps organizations of all sizes digitally transform using software, hardware and services from the Siemens Xcelerator business platform. Siemens’ software and the comprehensive digital twin enable companies to optimize their design, engineering and manufacturing processes to turn today’s ideas into the sustainable products of the future. From chips to entire systems, from product to process, across all industries. Siemens Digital Industries Software – Accelerating transformation.