Standardization News

“As artificial intelligence increasingly permeates every corner of life — from automated healthcare diagnoses to algorithm-driven hiring decisions — the need for robust regulatory frameworks has never been more urgent. In the United States, as tech giants and startups alike push the boundaries of what AI can do, lawmakers are attempting to keep pace. The risks of unregulated AI are stark: algorithms that reinforce bias, automated systems that invade privacy, and technologies that could reshape labor markets. With AI poised to revolutionize industries from finance to transportation, questions about how to govern this powerful tool are moving from the realm of theoretical debate to urgent policymaking.”

The above paragraph is the work of ChatGPT generated from the prompt: “Write the introduction of an article on the development of standards for artificial intelligence using New York Times style.”

In about seven seconds, the introduction was generated. Should journalists be concerned for their jobs? Not likely.

FOR YOU: ASTM's Aaron Prather Talks Robotics, AI, and Where the Industry is Going

While artificial intelligence (AI) and the closely related field of machine learning (ML) (the development of statistical algorithms that can learn from data to perform tasks without explicit instructions) offer a plethora of time-saving perks that help streamline processes and manage the mundane, it will most likely not take over the world and replace humans. However, the time is right to consider incorporating AI standards into processes and manufacturing.

The Launch of AI

While most people think of ChatGPT when they think of the birth of AI, the truth is, the concept of AI has been around much longer. Many consider British logician Alan Turing to be a key pioneer, having first suggested the concept in 1935. Turing went on to develop the “Imitation Game” that evaluated a machine’s ability to exhibit human-like intelligence. A few years later, Dartmouth College formalized AI as an academic discipline and hosted AI workshops that prompted labs around the globe to begin researching the possibility of AI. In the mid-1980s, AI became a billion-dollar industry, but research stagnated until the late 1990s, when technological advancements and improvements in computing speed led to a resurgence of research. By 2015, Google had more than 2,700 AI projects in development and numerous research projects were
in existence.

Standards help make AI safe and efficient.

On Nov. 30, 2022, ChatGPT was launched. In its first two months, the generative AI chatbot developed by OpenAI had 100 million users. Today, more than 180.5 million users ask ChatGPT to help them complete a variety of tasks, from writing essays about Ancient Rome, to churning out a biography on Marie Curie, to writing code for a computer-generated image, to asking ChatGPT to ask difficult questions to help prepare for a job interview.

Making Life More Efficient

While a 2023 survey from Pew Research reported that more than half (52%) of Americans are more concerned than excited about AI in daily life, there is no denying that AI does offer some benefits. For example, AI can be used to:

— Decrease human error. When AI-enabled computers are programmed correctly, they can run functions with zero errors.

—Automate repetitive tasks and improve processes. AI can be used to complete routine tasks, such as data entry and email responses, allowing employees to focus on more complex tasks.

— Manage big data. AI can process, analyze, and interpret big data in less time than humans.

— Act as a digital assistant. AI can collect reliable insights quickly and reduce decision-making time. Additionally, AI-based chatbots can be used to handle routine customer queries, thereby reducing the need for customer service.

— Perform dangerous tasks. AI robots can be used to engage in tasks that pose human risks, such as coal mining and rescue operations.

— Guide medical decisions and procedures. AI can help practitioners predict health risks quicker and assist with treatment and surgery.

— AI provides these benefits without requiring a vacation day, a lunch break, or time to sleep.

Standards Aim to Address Apprehension

Despite the benefits offered by AI, there is still some apprehension about fully embracing AI in daily life. According to Pew, Americans support the incorporation of AI into their daily lives to help reduce the effort and time spent on routine tasks. However, they also want regulations and oversight of emerging AI technologies.

READ MORE: The Next Industrial Revolution

To reap the benefits of AI while easing public apprehension, various ASTM International committees are investigating ways to incorporate AI into new standards. A recent ASTM internal survey on the use of AI and ML in committees revealed that the two are steadily making their way into the standards-development process and that committees are actively engaging the issue:

— 57% of respondents reported that their committees have discussed AI or machine learning.

— 30% of respondents have active work items on various AI topics, including AI, machine learning, deep learning, big data analytics, and smart assistants.

— 33% of respondents have published standards related to AI topics.

— 67% of respondents expressed an interest in participating in an ASTM forum to share information regarding the incorporation of AI into standards.

Shane Collins, member of the committee on additive manufacturing technologies (F42), sees AI having a big impact in his field. “The biggest impact from AI will be the speed at which novel system designs can be optimized and printed with additive-manufacturing technologies,” he says. “Organic, bone-like structures with only the exact amount of material necessary to achieve a superior design, and the lowest carbon footprint possible will combine to advance mobility, communications, medicine, and housing. AI-driven designs will need standards in order to ensure public safety as new approaches to manufacturing are realized with the combination of AI and AM.”

ASTM staff manager Molly Lynyak sees AI as providing an extra measure of safety and streamlining, with ASTM’s standards-development process being critical.

“The great thing about ASTM standards is that they are living documents that can be developed, revised, and changed as new technologies come to market, industries change and evolve, and new safety hazards are presented that need standardization to address those needs,” she says. “AI and ML will likely continue to evolve and advance in their capabilities and applications, and ASTM is ready to support the needs of our committees as they consider how to incorporate requirements and considerations for these technologies into their standards today and in the future.”

ASTM will continue to engage in discussions regarding AI and ML and identify ways to incorporate them into its standards to help bring these powerful tools to society and world.

Incorporating AI Into Standards

AI/ML is being incorporated into standards development across ASTM. Here are some of the most prominent examples:

Committee on consumer products (F15)

— F3463: Standard guide for ensuring the safety of connected consumer products

— F3698: Standard specification for computer-vision drowning detection systems for residential swimming pools

Committee on unmanned Aircraft Systems (F38)

— AC377 TR-1: Autonomy design and operations in aviation: terminology and requirements framework

— AC377 TR-2: Developmental pillars of Increased autonomy for aircraft systems

— AC377 TR-3: Regulatory barriers to autonomy in aviation white paper – roles and responsibilities for operational control in the age of increasingly autonomous flight

Committee on additive manufacturing technologies (F42)

— WK73978: Standard specification for additive manufacturing for metals – general principles – Registration of data acquired from process monitoring and for quality control

— WK76970: Standard guide for additive manufacturing – general principles – guidelines for technical and intellectual property authentication and protection

— WK78322: Standard guide for additive manufacturing – general principles – guidelines for AM security

— WK85641: Standard specification for additive manufacturing-data – data packages for AM parts

— WK89945: Standard specification for additive manufacturing – data – common data exchange format for laser blown powder directed energy deposition (DED)

— F3490: Standard practice for additive manufacturing – general principles – overview of data pedigree

— F3560: Standard specification for additive manufacturing – data – common exchange format for particle size analysis by light scattering

— F3605: Standard Guide for additive manufacturing of metals – data – file structure for inprocess monitoring of powder bed fusion (PBF). ●

Regan Daley is a communications intern with ASTM International.