AI-driven sustainability for small chemical manufacturers: Practical innovations for 2025 & beyond
Shehan Makani of ChemeNova and Chemrich Global outlines seven practical opportunities where sustainability aligns with cost reduction for small and mid-sized chemical manufacturers
Most sustainability programmes start with carbon emissions but, for small chemical manufacturers, material waste has a far larger impact on both cost and environmental footprint. Studies at the New Jersey Institute of Technology( NJIT) show that every kilogram of wasted speciality chemical carries five to 20 times more embedded energy and CO 2 than the equivalent emissions from equipment operation.
Waste is the highest leverage point here for several reasons. It means a loss of money; costs in energy, transport, processing and disposal; and the traceability burden under evolving global regulations. In pilot work conducted via ChemeNova’ s IntelliForm * platform and Chemrich Global’ s small-batch manufacturing network, firms saw up to 12 % cost savings simply by analysing‘ hidden waste’ originating from over-ordering, shelf-life expiry and inaccurate batch transfer.
Chemical SMEs often over-purchase raw materials due to high minimum order quantities( MOQs). This creates cash-flow strain and avoidable waste. The emerging solution to this is AIdriven split-lot procurement( Figures 1 & 2). Using lightweight large language models( LLMs), ChemeNova has developed a prototype that:
• Matches multiple buyers needing the same material
• Predicts optimal pooling combinations
Figure 1 – AI-enabled sustainable chemical manufacturing
• Verifies safety data sheet( SDS) and certificate of analysis( CoA) data and supplier claims
• Suggests safer or greener substitutes
This‘ micro-procurement’ model reduces waste before it ever occurs and addresses a major market gap where large marketplaces favour bulk buyers, over small manufacturers.
Supplier selection is still a manual, time-intensive process for SMEs. My recent research has highlighted how SMEs lack access to data-driven sourcing tools used by global firms. 1 AI can close that gap with:
• Optical character recognition( OCR) and LLM extraction from SDS and COA
• Automated hazard flagging
• Verification of biobased, recycled or low-VOC attributes
• Cross-matching cost, MOQs and sustainability metrics
Chemrich Global has begun embedding this into its speciality sourcing operations, enabling smaller customers to adopt greener suppliers without adding workload.
Yield loss is the single most expensive inefficiency in chemical manufacturing. Yield analytics remains one of the easiest, most cost-effective AI upgrades for small manufacturers. Low-cost digital yield analytics can detect multiple problems, notably:
• Operator variability
• Raw material inconsistencies
• Temperature or humidity deviations
• Improper cleaning
• Inefficient mixing parameters Using a combination of simple scales, BLE sensors, and a Pythonbased model initially developed during NJIT’ s Smart Manufacturing showcase, companies achieved a 3 – 10 % yield improvement,
56 SPECIALITY CHEMICALS MAGAZINE ESTABLISHED 1981