SUPPLY CHAIN MANAGEMENT the system to estimate impacted deliveries , suggest alternative warehouse stock or sources of new supply outside the area likely to be affected , and run through the transport options as noted above , quickly and through a single intuitive interface .
AI-based supplier selection
Another vital aspect of supply chain management is in supplier selection and relationship management . That is about understanding who your suppliers are , what they are doing for you and how well they are doing it .
AI can evaluate suppliers based on criteria such as cost , quality and reliability . It can take into account risk factors , such as production facilities in hurricane zones and help to establish portfolios of supply alternatives . That is crucial in an era where dependence on particular vendors , or even particular regions , can represent unacceptable business risk .
Let ’ s move way up the supply chain to the two aforementioned areas of AI that are , based on my discussions around the industry , of principal interest to speciality chemicals producers : AI product development and AI predictive quality .
AI product development
AI product development is fast becoming an industry standard , sparing untold hours of speculative lab work through predictive modelling of new recipes ; formulation optimisation to improve effectiveness and stability ; molecular-level simulations of reactions and processes ; and process optimisation to boost yield , cut waste , and recognise safety and regulatory risks that may be associated with new compounds .
Integration with business systems can add to the ( conceptual ) mix the commercial success of similar formulations , among other variables . From the perspective of the business , AI product design boils down to adapting products to meet changing market demands , which enhances supply chain stability by meeting market needs and thereby stoking demand for the supply chain to fulfil . At the same time , turbocharging the pace of product innovation with AI will put immediate pressure on the supply chain . There will be new inputs for existing suppliers to deliver ; new suppliers to vet and bring into the fold quickly ; new raw materials to source ; new safety and regulatory conditions to manage ; and new production , storage and transport approaches to model and realise .
It will have to happen at an unprecedented pace that will all but demand supply chain AI capabilities . AI product design is shaping up to be among the greatest business catalysts this industry so deeply familiar with actual catalysts has ever seen .
AI predictive quality
AI predictive quality — a close runnerup to AI product development in terms of the speciality chemicals industry ’ s AI priorities at the moment — also has serious supply chain implications . It takes aim at the outputs of AI product design , the goal being the consistent production of ‘ golden batches ’ for those new products . ( These systems also work with existing products and processes derived from biological ingenuity , of course .)
AI predictive quality simulations take into account technical data , such as raw material properties , processing temperature and reaction times , as well as business data , such as facts related to the supplier of the raw material , to help monitor and adjust production processes to optimise every batch . The importance of predictive quality for the supply chain lies in the fact that it avoids unplanned delivery problems due to poor quality or recognises them at an early stage , so that countermeasures can be taken as quickly and effectively as possible .
Those considerations do not end at the walls of the production facilities . The quality of the product depends on the quality of its inputs . Rarely are the compositions of production inputs perfectly consistent with their nominal
specifications ; even ‘ pure ’ substances have their impurities .
AI predictive quality systems can analyse the performance of inputs from different suppliers and establish how certain inputs affect end-product quality — information you can feed back to suppliers ( and they can feed back to their suppliers ) to boost quality down the chain .
Outlook
Predicting AI systems ’ ultimate influence on the speciality chemicals industry is no easier than predicting hurricanes months in advance . We know what realms are the likely targets and we know that they will have a major impact sooner than later .
Where the hurricane of AI differs is that , where an actual hurricane ’ s influence is spatially and temporally limited , AI capabilities will ultimately touch the entirety of a speciality chemicals business , and they will combine operational and business data to help make better decisions quickly .
All in all , AI is shaping up to be a hurricane unto itself . The good news is that this particular hurricane looks to be an overwhelmingly positive force for change . ●
J j
Dr Marko Lange
HEAD OF CHEMICALS
SAP marko . lange @ sap . com www . sap . com
JUL / AUG 2024 SPECCHEMONLINE . COM
43