AI is infiltrating engineering software, from 3D design to software development and beyond.
Engineers can leverage a rapidly growing list of generative AI software services to operate with a data-driven management philosophy. Some of the same software can enhance, accelerate and cost-contain their digital transformation initiatives.
Digital transformation aims to move organizations to a data-driven management culture that increases data-driven decision-making. This culture reduces operational risk, increases revenue and reduces costs.
Most software vendors are adding generative AI features to their software and services. Examples include well-established vendors, such as Google, IBM, Microsoft, Oracle, and SAP, as well as new entrants, such as Databricks, Dataiku, Open AI and Snowflake. These AI features add value for engineers working on the following aspects of digital transformation.
Data management
All digital transformation advances and data-driven decision-making depend on data management improvements such as more systems integration, data lakehouses, data warehouses and data restructuring. The software vendors supplying these capabilities are rapidly incorporating generative AI features into their products. For example, Oracle Autonomous Database Select AI integrates large language models (LLMs) with generative AI data management applications.
Data management includes acquiring, validating, storing, protecting and processing organizational and external data. The goal is to provide engineers and other disciplines with timely access to comprehensive, integrated data for analysis and data-driven decision-making.
Widely implemented data management software includes AWS Redshift and EMR, Databricks, Dataiku, Google Cloud Platform, Microsoft Azure Data Lake Store and Synapse Analytics, Oracle Exadata Cloud Service, SAP HANA Cloud and Snowflake.
Data quality
The leading risk engineers experience during digital transformation initiatives is inadequate data quality. The software vendors that help organizations improve their data quality processes are rapidly incorporating generative AI features into their products. For example, Informatica Data Quality applies AI-driven models to recognize many invalid or missing values and correct these to improve data quality.
Data quality management includes data preparation, enrichment and quality monitoring. The value is to assure end-users that data is accurate, consistent and reliable for data-driven decision-making.
Widely implemented AI-assisted data quality software includes Alteryx, IBM InfoSphere, Informatica Data Quality, Oracle Enterprise Data Quality (EDQ), Precisely Trillium, SAP Data Intelligence, SAS Data Quality and Talend Data Fabric.
Software development
Many digital transformation projects include some custom software development. Handcrafting software is expensive and subject to annoying defects.
Many AI-enhanced coding assistants have emerged that support various programming languages and integrated development environments (IDEs). For example, Amazon CodeWhisperer uses AI to suggest source code ranging in size from snippets to complete functions. The benefit of coding assistants is significantly improving software development productivity, which hasn’t improved much in many years.
Widely implemented AI software development products include Amazon CodeWhisperer, Blackbox, CodiumAI Codiumate, GitHub Copilot, Google AlphaCode 2, Google Cloud’s Duet and Tabnine. Their capability, use of AI and number of supported programming languages vary considerably.
Software unit testing
Software unit testing is a tedious task that’s an inescapable part of software development for digital transformation. Unit tests verify that the software works as intended, catch defects early and ensure that code is maintainable.
For example, Diffblue Cover uses AI reinforcement learning to create accurate and maintainable unit tests more efficiently than manual test development. The benefit of automated unit testing tools is significantly improved software testing coverage. Many software projects shortcut testing due to the pressure to complete projects.
Widely implemented AI-assisted automated unit testing tools include Bito, DeepUnitAI, Diffblue Cover, EvoSuite, Functionize, Mabl, Symflower, Testim, Squaretest and Testsigma.com. These AI-assisted unit testing tools vary in degree of automation.
3D design
Generative AI software assists engineers in 3D design by exploring a vast design space, identifying novel solutions, and testing them. It sometimes generates unexpected design alternatives, leading to more efficient, innovative designs and data-driven decision-making.
Engineers benefit from rapidly exploring innovative product designs that can be developed or manufactured. For example, Blender supports the AI art generator Stable Diffusion through Stability for Blender, a Stability AI text-to-image generator add-on.
Widely implemented design software includes Blender, Catia, Creo, Fusion 360, Solid Edge and Solidworks.
Content creation
Engineers working on digital transformation and data-driven decision-making will produce lots of content. Examples include manuals, project deliverables like requirements and designs, ongoing reports to management and websites. They’ll also perform data analysis and summarize lengthy documents created by others.
Generative AI software understands and responds in contextually relevant ways to content requests while supercharging the productivity of engineers. The benefit of AI for content creation is lower effort for text.
Widely implemented AI content creation software includes ChatGPT-4 from Open AI, Google Gemini (which replaced Bard), Claude 3 from Anthropic, Microsoft Copilot (which replaced Bing AI) and Poe.
Challenges still arise due to the potential for misinformation, bias and limitations in comprehending intricate contextual engineering nuances.
Research
Engineers frequently research specific topics to understand technical developments or to diagnose manufacturing problems. AI-driven generative software for research combines the strengths of ChatGPT for a summary of the information on a topic and a search engine for web links engineers can follow to explore more detailed information.
For example, Perplexity is an AI-chatbot-powered research and conversational search engine that answers queries using natural language predictive text. It provides sources with links in its responses.
The benefit of research software for data-driven decision-making is the assurance that most potentially useful sources have been included in the result set. This assurance is difficult to achieve with more manual research.
Widely implemented AI research software includes Consensus, Liner and Perplexity.
Breaking through the barriers of generative AI
Applying generative AI to digital transformation can feel both daunting and risky. The common concerns that make engineering leaders who are reluctant to adopt generative AI for data-driven decision-making include the following:
- Inadvertently introducing security and data privacy risks
- System integration challenges
- Surprisingly high development and operating costs
- Lock-in and technical dependencies on specific vendors and products
- Lack of specialized AI software talent
- Lack of auditability
- Insufficient regulatory compliance
- Lack of end-user trust in the reliability of generative AI work products
- Lack of control over ungoverned shadow AI usage
Actions that can reduce these inhibitors to digital transformation include starting with specific applications for a single department, such as engineering design or a business function, such as supply chain. Successful digital transformation avoids an enterprise-wide rollout. Focusing on business value rather than technological wizardry is more likely to lead to the successful adoption of data-driven decision-making.
Generative AI has ushered in a new era of possibilities, where software mimics the creative processes of engineers and actively contributes to digital transformation. From generating code snippets to improving engineering data access to crafting visual content, AI software demonstrates the remarkable progress of technology to support the data-driven management philosophy.
Engineers recognize both the potential and limitations of generative AI. They increasingly embrace its power while understanding the need for oversight and refinement. Engineers who explore these cutting-edge generative AI tools are well on their way to achieving unprecedented efficiency, creativity and innovation.
We can expect these AI tools to advance in functionality rapidly in the near term, continuing to improve engineering productivity and quality for digital transformation and data-driven decision-making.
For more, check out 6 ways engineers can use generative AI in 2024.
—
Yogi Schulz has over 40 years of Information Technology experience in various industries. He writes for IT World Canada and other trade publications. Yogi works extensively in the petroleum industry to select and implement financial, production revenue accounting, land & contracts, and geotechnical systems. He manages projects that arise from changes in business requirements, from the need to leverage technology opportunities and from mergers. His specialties include IT strategy, web strategy, and systems project management.
