Back to Resources
Blog

Using AI to Create the Next Generation of Cleaning Products

Written by
January 14, 2025
Share this post
Copy Link

Under almost every kitchen sink in the developed world, you will likely find an array of different cleaning products poised and ready to be called into action on a moment's notice. These liquids, sprays, gels, and powders are miraculous: removing grime, grease, stains, bacteria, viruses and more from surfaces you were convinced may never be clean again. It is undeniable that cleaning products have improved our standard of living tremendously: mitigating the very real and immediate threats of infectious disease and reducing the toil of household and industrial labor.

However beneficial, these products also unfortunately present their own hazards to human health. As an example, occupational exposure to cleaning products and disinfectants puts professional cleaners in domestic or hospital settings at risk of chronic obstructive pulmonary disease, increased rates of death due to chronic obstructive pulmonary disease (COPD), and accelerated decline in lung functions1. And even household cleaners are connected with increased risks of asthma, accelerated lung function decline, allergen sensitization, and cancer1. As the long-term effects of hazardous ingredients are realized, there has been increasing demand for regulation and development of safer formulations. This is evidenced by the recent EPA ban2 on perchloroethylene, a dry cleaning agent found to cause several types of cancer and cause damage to multiple organs and systems in the body3. Therefore, a preeminent challenge for the cleaning industry today is to create the next generation of products with improved safety that comply with an evolving regulatory landscape.

On the surface level, it might seem that creating a ‘safer’ product would simply mean replacing a ‘dangerous’ chemical with one that is less-so. Simple, right? However, cleaning products are complex technologies with many considerations. Active ingredients can be especially hard to replace because the very properties that make them effective—like acidic or caustic elements that dissolve minerals and grime—are inherently hazardous. While proper safety practices help, user exposure is inevitable1. This creates a fundamental tension between product effectiveness and user safety. Non-active ingredients also present their own challenges. Common additives like EDTA (a water softener) and fragrances can pose health risks4,5, yet they serve important functions in product stability, performance, and user experience. Beyond this, even when better ingredients are found, formulations need to be adjusted and optimized to ensure that key properties are maintained.

The cleaning industry therefore faces complex questions about balancing short-term cleaning effectiveness with long-term health impacts, particularly in both consumer and industrial applications. These challenges intersect with regulatory pressures, consumer demands, and corporate responsibilities, requiring careful consideration of how to maintain product efficacy while improving safety. In the face of such multifaceted and nuanced factors, what strategies can the cleaning industry employ to accelerate the pursuit of safer and more compliant products?

The reformulation challenge

Keeping up with the rapidly evolving regulatory environment can be challenging, especially for companies who distribute products internationally and are subject to oversight by multiple regulatory bodies. Formulas that were previously regarded as acceptable may have ingredients that will be restricted in the future due to expanding safety requirements. There are therefore two primary objectives for companies creating their next generation of products. First, they must define the safety metrics and goals they’d like for their new products. These goals must be a balance between realistic and achievable improvements to ensure short-term profitability, and ideally would lead to formulations that can last as long as possible in the face of ever-updating regulations. Second, they must find a way to replace ingredients that are at-risk of regulation with ones that are safer and therefore not likely to be regulated soon.

While companies may be aware of ingredients that are on the near-term regulatory chopping-block, there are still many questions to answer and obstacles to overcome in the development of new products. The challenge here is managing the complex array of ingredients and properties, and having a systematic and precise way to find new ingredients for each of the desired properties. How do companies navigate the astronomically large combinatorial space of potential formulations under more stringent constraints in a rapid and cost effective manner? Are there ways to predict the safety of potential replacement ingredients? Though daunting, the good news is that new Science Based Artificial Intelligence (SBAI) techniques have emerged to provide answers and solutions to these forthcoming questions and problems.

How SBAI can advance reformulation

Predicting chemical properties is a difficult task. However, the field of chemistry has enjoyed advances commensurate with the magnitude of those in computation! More specifically, the development of machine learning (ML) and AI technologies has enabled incredible innovation for overcoming complex and multi-dimensional problems like those faced in chemical development. No longer do we need to painstakingly measure and slowly whittle down every possible formula. Instead, machine learning researchers can discover unique insights by leveraging pre-existing data before product developers need to return to the lab, saving time and money. What this means for the cleaning industry is that ML algorithms can help them both identify new ingredients with similar properties while avoiding undesirable side-effects and also even help optimize the ratios of ingredients to maximize performance and cost. Moreover, the use of Science-Based AI (SBAI) can be even more powerful than traditional AI techniques alone. By infusing scientific expertise into models, chemical and formula space can be navigated with much more precision, meaning less data is needed to train effective predictive algorithms, and little to no time is wasted exploring undesirable options.

NobleAI’s Risk Assessment and Ingredient Replacement (RAIR) solution

At NobleAI, we are leading experts in SBAI for chemistry and material science applications. Our Risk Assessment (RA) services incorporate the latest restricted substance lists across the globe to help you develop products for international distribution, and our chemical structure similarity algorithms can help anticipate potential chemical concerns by matching compounds in your formulas with high similarity to known hazards via structural similarity assessments. Whether you are looking to evaluate your current portfolio or working on your next generation of products, NobleAI’s SBAI and Visualization, Insights, and Predictions (VIP) Platform can help accelerate your product development in minutes, not months.

Our SBAI models can take into account multiple company-specific safety goals and performance requirements simultaneously to provide bespoke insights tailored for your direct needs. Work with our AI team and domain experts to reformulate your product with our Ingredient Replacement (IR) capabilities and optimize your formulations further. To learn more about our RAIR solution, or other solutions offered on the VIP Platform: schedule a discovery call, visit our website, or contact us today!

NobleAI Risk Assessment Results of Editic Acid (EDTA)

NobleAI Risk Assessment Results Continued: Similarity to Restricted Substances

Citations:
  1. Salonen, Heidi, et al. "Cleaning products: Their chemistry, effects on indoor air quality, and implications for human health." Environment International (2024): 108836.
  2. https://www.epa.gov/newsreleases/biden-harris-administration-announces-latest-actions-under-nations-chemical-safety-law
  3. Ceballos, Diana M., et al. "Perchloroethylene and dry cleaning: It's time to move the industry to safer alternatives." Frontiers in Public Health 9 (2021): 638082.
  4. Long, David C. "Greening of consumer cleaning products." Green Techniques for Organic Synthesis and Medicinal Chemistry (2018): 91-115.
  5. Steinemann, Anne. "Fragranced consumer products: exposures and effects from emissions." Air Quality, Atmosphere & Health 9.8 (2016): 861-866.

Using AI to Create the Next Generation of Cleaning Products

Written by
January 14, 2025
Share this post

Under almost every kitchen sink in the developed world, you will likely find an array of different cleaning products poised and ready to be called into action on a moment's notice. These liquids, sprays, gels, and powders are miraculous: removing grime, grease, stains, bacteria, viruses and more from surfaces you were convinced may never be clean again. It is undeniable that cleaning products have improved our standard of living tremendously: mitigating the very real and immediate threats of infectious disease and reducing the toil of household and industrial labor.

However beneficial, these products also unfortunately present their own hazards to human health. As an example, occupational exposure to cleaning products and disinfectants puts professional cleaners in domestic or hospital settings at risk of chronic obstructive pulmonary disease, increased rates of death due to chronic obstructive pulmonary disease (COPD), and accelerated decline in lung functions1. And even household cleaners are connected with increased risks of asthma, accelerated lung function decline, allergen sensitization, and cancer1. As the long-term effects of hazardous ingredients are realized, there has been increasing demand for regulation and development of safer formulations. This is evidenced by the recent EPA ban2 on perchloroethylene, a dry cleaning agent found to cause several types of cancer and cause damage to multiple organs and systems in the body3. Therefore, a preeminent challenge for the cleaning industry today is to create the next generation of products with improved safety that comply with an evolving regulatory landscape.

On the surface level, it might seem that creating a ‘safer’ product would simply mean replacing a ‘dangerous’ chemical with one that is less-so. Simple, right? However, cleaning products are complex technologies with many considerations. Active ingredients can be especially hard to replace because the very properties that make them effective—like acidic or caustic elements that dissolve minerals and grime—are inherently hazardous. While proper safety practices help, user exposure is inevitable1. This creates a fundamental tension between product effectiveness and user safety. Non-active ingredients also present their own challenges. Common additives like EDTA (a water softener) and fragrances can pose health risks4,5, yet they serve important functions in product stability, performance, and user experience. Beyond this, even when better ingredients are found, formulations need to be adjusted and optimized to ensure that key properties are maintained.

The cleaning industry therefore faces complex questions about balancing short-term cleaning effectiveness with long-term health impacts, particularly in both consumer and industrial applications. These challenges intersect with regulatory pressures, consumer demands, and corporate responsibilities, requiring careful consideration of how to maintain product efficacy while improving safety. In the face of such multifaceted and nuanced factors, what strategies can the cleaning industry employ to accelerate the pursuit of safer and more compliant products?

The reformulation challenge

Keeping up with the rapidly evolving regulatory environment can be challenging, especially for companies who distribute products internationally and are subject to oversight by multiple regulatory bodies. Formulas that were previously regarded as acceptable may have ingredients that will be restricted in the future due to expanding safety requirements. There are therefore two primary objectives for companies creating their next generation of products. First, they must define the safety metrics and goals they’d like for their new products. These goals must be a balance between realistic and achievable improvements to ensure short-term profitability, and ideally would lead to formulations that can last as long as possible in the face of ever-updating regulations. Second, they must find a way to replace ingredients that are at-risk of regulation with ones that are safer and therefore not likely to be regulated soon.

While companies may be aware of ingredients that are on the near-term regulatory chopping-block, there are still many questions to answer and obstacles to overcome in the development of new products. The challenge here is managing the complex array of ingredients and properties, and having a systematic and precise way to find new ingredients for each of the desired properties. How do companies navigate the astronomically large combinatorial space of potential formulations under more stringent constraints in a rapid and cost effective manner? Are there ways to predict the safety of potential replacement ingredients? Though daunting, the good news is that new Science Based Artificial Intelligence (SBAI) techniques have emerged to provide answers and solutions to these forthcoming questions and problems.

How SBAI can advance reformulation

Predicting chemical properties is a difficult task. However, the field of chemistry has enjoyed advances commensurate with the magnitude of those in computation! More specifically, the development of machine learning (ML) and AI technologies has enabled incredible innovation for overcoming complex and multi-dimensional problems like those faced in chemical development. No longer do we need to painstakingly measure and slowly whittle down every possible formula. Instead, machine learning researchers can discover unique insights by leveraging pre-existing data before product developers need to return to the lab, saving time and money. What this means for the cleaning industry is that ML algorithms can help them both identify new ingredients with similar properties while avoiding undesirable side-effects and also even help optimize the ratios of ingredients to maximize performance and cost. Moreover, the use of Science-Based AI (SBAI) can be even more powerful than traditional AI techniques alone. By infusing scientific expertise into models, chemical and formula space can be navigated with much more precision, meaning less data is needed to train effective predictive algorithms, and little to no time is wasted exploring undesirable options.

NobleAI’s Risk Assessment and Ingredient Replacement (RAIR) solution

At NobleAI, we are leading experts in SBAI for chemistry and material science applications. Our Risk Assessment (RA) services incorporate the latest restricted substance lists across the globe to help you develop products for international distribution, and our chemical structure similarity algorithms can help anticipate potential chemical concerns by matching compounds in your formulas with high similarity to known hazards via structural similarity assessments. Whether you are looking to evaluate your current portfolio or working on your next generation of products, NobleAI’s SBAI and Visualization, Insights, and Predictions (VIP) Platform can help accelerate your product development in minutes, not months.

Our SBAI models can take into account multiple company-specific safety goals and performance requirements simultaneously to provide bespoke insights tailored for your direct needs. Work with our AI team and domain experts to reformulate your product with our Ingredient Replacement (IR) capabilities and optimize your formulations further. To learn more about our RAIR solution, or other solutions offered on the VIP Platform: schedule a discovery call, visit our website, or contact us today!

NobleAI Risk Assessment Results of Editic Acid (EDTA)

NobleAI Risk Assessment Results Continued: Similarity to Restricted Substances

Citations:
  1. Salonen, Heidi, et al. "Cleaning products: Their chemistry, effects on indoor air quality, and implications for human health." Environment International (2024): 108836.
  2. https://www.epa.gov/newsreleases/biden-harris-administration-announces-latest-actions-under-nations-chemical-safety-law
  3. Ceballos, Diana M., et al. "Perchloroethylene and dry cleaning: It's time to move the industry to safer alternatives." Frontiers in Public Health 9 (2021): 638082.
  4. Long, David C. "Greening of consumer cleaning products." Green Techniques for Organic Synthesis and Medicinal Chemistry (2018): 91-115.
  5. Steinemann, Anne. "Fragranced consumer products: exposures and effects from emissions." Air Quality, Atmosphere & Health 9.8 (2016): 861-866.
Sign Up for Updates