How AI Can Potentially Fix Some of Sustainable Fashion’s Nagging Challenges

From helping choose more sustainable materials to optimizing production and automated sorting, AI is offering more creative ways to solve fast fashion’s sustainability challenges. 

Sustainability is more often talked about than practised. A key reason is that sustainability has some hard challenges. And brands often find it overwhelming to address those with the solutions they have. 

Think of this: each year, the global fashion industry produces over 100 billion garments, out of which 92 million tonnes end up in landfills. A substantial portion gets burned as well. Also, less than 1% of the material used to produce clothing is recycled into new clothing.

The recycling numbers could be much higher if sorting weren’t that difficult. The garments we wear are not made of homogeneous materials: there are different types of textiles, fibres, and fixtures. 

It’s immensely resource-intensive to sort these into different categories before recycling them. Artificial Intelligence changes this with computer vision that can sort materials into different categories without human intervention. 

Modern problems need modern solutions - or, so goes the saying. Here are five ways AI is set to solve some of the hard problems of sustainability in fashion: 

1) Cutting overproduction through better demand forecasting 

Of the 100 billion garments produced annually, roughly 20-30%, which makes it about 30 billion garments, go unused. No one ever wears them. So why do companies produce so much even though they are unable to sell everything?

A part of the problem is poor demand forecasting.

Fashion trends change fast, and companies often find it difficult to reliably know which styles are going to sell more and by how much. We cannot effectively solve this problem through market surveys and traditional tools of demand forecasting, which are largely backward-facing and rely heavily on historical data.

AI-powered tools are more effective at analysing real-time signals such as what people search for online, how they talk about fashion trends on social media, and how consumer sentiment shifts across regions and demographics. 

By combining historical sales data with real-time behavioural signals, AI models can generate more accurate demand forecasts at the SKU and regional level.

This helps brands to produce closer to actual market demand, cut inventory, and limit the number of garments that eventually end up being dumped or burned. 

For example, AI-powered platform Stylumia analyzes real-time data from online searches, social media trends, and historical sales numbers to forecast demand. 

2) Virtual try-on and AI sizing to reduce returns (and emissions)

Return rates in fashion products are much higher than in other categories like food & beverages and electronic gadgets. According to a Statista survey, more than 50% of US consumers have reported returning a garment, shoe, or accessory that they bought online. 

Source: Statista 

Frequent returns not only jack up supply chain costs but also lead to a higher carbon footprint. Some of the key reasons people return clothes or shoes include fitting issues and a change of opinion after trying them at home. Brands are increasingly realizing that making hyper realistic virtual trial room and sizing can cut returns significantly. 

AI-powered virtual try-on, 3D fit prediction, and AI sizing help in recommending the best size while helping shoppers visualize their best fit. These tools can blur the distinction between online shopping and shopping at physical stores. Studies show that an effective virtual try-on can lower return rate by 20-30%. 

Google recently launched an AI-powered tool that allows users to select a style, upload their full-body photo, and look at how it actually looks on them before making a purchase decision. 

Source: blog.google

3) AI-assisted design and pattern optimization

An enormous amount of fabric gets wasted long before garments even reach a store. Studies show that up to 25% of fabric can be wasted during the design, garment testing, and pattern modification stage alone.

So, every prototype that doesn’t eventually end up in stores counts as a waste. AI-assisted design and pattern optimization can lower fabric waste. 

Tools like Six Atomic, which integrates directly into design software, reduce design time from hours to minutes. They can also send patterns straight to manufacturers, often getting the physical sample right on the first attempt. 

In addition to prototyping, AI tools can also optimize how pattern pieces are laid out on a fabric roll to ensure that the maximum amount of material is used. Brands using AI-powered cutting optimization have reported up to a 20% improvement in fabric utilization efficiency and a 30% reduction in sampling waste. 

4) Automated textile sorting and recycling

While we touched on this point at the beginning of this blog, it’s worth digging deeper because sorting and recycling can immensely benefit from AI applications.

What makes physical sorting really difficult is that a typical garment isn't made of one material. It's a blend of cotton, polyester, and spandex, with buttons, zippers, dyes, and stitching thrown in. Sorting these by hand, at scale, is practically impossible.

AI-powered computer vision and hyperspectral imaging are addressing these challenges. US-based startup Refiberd uses AI paired with hyperspectral cameras to detect fiber composition in textile waste with high accuracy: 

Source: Refiberd

Their technology aims to divert up to 70% of textile waste to high-value recyclers. 

5) Supply-chain traceability and factory efficiency 

Supply chains in the fast fashion industry are notorious for their lack of transparency. So, while a garment may be sold in a flashy store in New York, its cotton may have been farmed in India, spun in Bangladesh, dyed in Vietnam, and assembled in Cambodia. 

At each step, there is water usage, carbon emissions, working conditions, land usage, and material sourcing decisions that brands have historically had little visibility into. 

AI is changing that by aggregating and analysing supply chain data in real time. This helps in flagging inefficiencies and helping brands make more responsible sourcing decisions. 

Platforms like Fairly Made are helping brands trace every step from raw material to finished product. These types of AI tools source data from suppliers, logistics partners, and third-party auditors, pulling in everything from factory energy consumption to shipping routes. 

Next, they generate a single picture of the entire supply chain. When brands can see exactly where emissions and inefficiencies are taking place, they can do something about it rather than staying in the dark.

Sustainability in fashion has always had more talkers than doers, and there are several myths that keep it that way. AI enables everyone to take proactive actions and do something about fashion’s nagging environmental challenges.