Unilever Engineers Detergent Formulas for Smart Washing Machines Using AI
Unilever has released two laundry products designed specifically for auto-dose washing machines: Persil Advanced Clean Non-Bio and Comfort Smart Series Azure Bliss. The detergent and conditioner were engineered using AI to function reliably inside automated dispensing systems rather than being manually poured.
The company developed the formulas using PRO-Cision technology, which models thousands of simulated scenarios based on consumer usage patterns, machine behavior, performance testing and formulation variables. The goal was identifying the precise fluid dynamics required for products to flow through auto-dose reservoirs without clogging or degrading.
The Technical Problem
Most existing detergents were formulated for manual dosing. Their viscosity and stability characteristics don't match the narrower pipes and reservoir systems in smart machines.
"Using a regular liquid in an automated machine is a little bit like using a thick syrup in an espresso machine," said Andrea Parra-Vera, a Unilever home care R&D specialist. "The espresso machine is built for precision liquid and the right flow, not a liquid that thickens in the tank."
When conventional liquid sits in a heated reservoir across multiple wash cycles, it can thicken, leave residue and cause blockages that degrade machine performance over time.
What Changed in the Formula
The Smart Series formula was engineered to remain stable under heating and repeated use, flowing cleanly through dosing mechanisms without leaving buildup while delivering stain removal at the precise doses set by the machine.
Fragrance stability presented a separate challenge. Auto-dose machines hold product in reservoirs for extended periods and subject it to repeated heating cycles, which can degrade volatile fragrance compounds before they reach the wash. Unilever redesigned the fragrance system to remain chemically stable under those stresses.
Samsung Partnership Shaped Development
Unilever collaborated with Samsung, whose washing machine performance and cycle data informed how the formula behaved under real-world conditions. Samsung's AI wash systems use sensors to assess load weight, fabric type and soil level, then auto-dispense detergent accordingly.
The partnership was technical as well as commercial. Machine cycle data directly shaped both the formulation and the packaging, which was redesigned to reduce spillage during reservoir refilling.
Market Timing
Auto-dose washing machines currently account for around 14% of U.S. households. Unilever projects that figure will rise to approximately 20% by 2030.
The Smart Series range will initially be available exclusively in the U.K.
For product development teams, this work illustrates how AI for Product Development can solve hardware-software compatibility problems. It also demonstrates how AI Agents & Automation systems require formulations engineered specifically for their constraints rather than adapted from existing products.
Your membership also unlocks:
