Artificial intelligence
In forecasting and optimizing pharmaceutical industry inventories
The pharmaceutical industry supply chain includes raw material and material suppliers, pharmaceutical manufacturers, distributors and retail chains. Each link in the above chain works together to provide consumers with products that can meet their quality and budget needs. Competition in the market is fierce, and succeeding in the market requires effective management of many processes.
Demand forecasting process
Challenges
Predictions of future demand volumes are the base on which all other processes in the company need to be based. Mistakes made at this stage amplify in all other activities of the company leading to inefficient allocation of capital resulting in loss of sales, margin, increased costs and reduced liquidity.
Obtaining high-quality forecasts is not an easy task. The average sales model approach, which is often used in practice, usually leads to out-of-stock as well as overstock situations. Lack of availability of a commodity causes customers to temporarily switch to buying competing goods, and in the long run they may change their product preference altogether. Excess inventory, in turn, means the risk of out-of-stock and the need to sell out eroding margins or incurring pharmaceutical disposal losses.
Rozwiązanie
Offering customers hundreds of products that fulfill their different needs requires a personalized approach to analyzing their behavior. In other words, it is necessary to adjust demand forecasting models from the perspective of each product in the context of individual customers, channels or locations. You have to approach forecasting seasonal products differently, event products differently, and year-round products differently. In addition, in assessing the volume of demand, it is necessary to take into account both demand-stimulating instruments that depend on us, such as promotions, and elements beyond our control, such as the weather.
Inventory optimization process
Challenges
Predictions of future demand volumes are the base on which all other processes in the company need to be based. Mistakes made at this stage amplify in all other activities of the company leading to inefficient allocation of capital resulting in loss of sales, margin, increased costs and reduced liquidity.
Obtaining high-quality forecasts is not an easy task. The average sales model approach, which is often used in practice, usually leads to out-of-stock as well as overstock situations. Lack of availability of a commodity causes customers to temporarily switch to buying competing goods, and in the long run they may change their product preference altogether. Excess inventory, in turn, means the risk of out-of-stock and the need to sell out eroding margins or incurring pharmaceutical disposal losses.
Solution
Using AI and ML in the process of generating demand forecasts and simulating inventory across the distribution network allows for a holistic view of the stock situation and an assessment of current and future out-of-stock and overstock risks. This makes it possible to undertake quick and flexible responses to changes.
The idea of exchanging information between a customer (e.g., a wholesaler) and a supplier (e.g., a manufacturer) necessary for joint forecasting and management of total inventory carries a number of benefits. However, it requires close cooperation in sharing information that is important to the overall process. Coordination of sales service is better, there are no surprises when planning inventory levels, e.g., for promotional activities, the average total inventory level is lower, transportation costs and environmental burden are lower.
Within the VMR model, several basic methods of managing the customer's mazana stocking can be listed, including: MIN-MAX and Continuous review with maintenance of the required Service Level.
Coupling the AI and ML-based forecasting process with inventory control and generating stocking recommendations at both the customer and supplier warehouse level gives a great advantage at the operational level. Decisions about what to produce, where to produce it and in what quantities to keep in-house or in customers' warehouses must be made every day for hundreds or thousands of products. Supporting this process with precise calculations of AI and ML algorithms and automation of data processing is also a huge increase in employee productivity. This is because they are relieved from performing repetitive operations of preparing and processing batch data and performing recalculations on their own. This gives you the opportunity to focus on the important aspects of your work: risk analysis and prevention, process control, dealing with suppliers, etc.
The Dature system uses artificial intelligence (AI) and machine learning (ML) to generate accurate demand forecasts and optimize inventory. The CPFR (Collaborative Planning, Forecasting and Replenishment) and VMR (Vendor Managed Replenishment) mechanisms used in the application for individualized approach to forecasting of individual SKUs, taking into account both internal organization data (sales, own promotions, etc.) and data from the environment (e.g. customer and supplier promotions, weather, etc.). This significantly improves the quality of forecasts and reduces the risk of out-of-stock and overstock situations. Thanks to the simulation and inventory optimization mechanisms available in the system, there is a systematic improvement in the quality of the entire process. This, in turn, carries over to financial effects: sales, margin, costs, liquidity, as well as image. In other words, much higher market competitiveness is achieved.
Potential business benefits
Reduce forecasting errors by 30% to 50%.
Reduce lost sales by up to 65%,
Reduction in transportation and administration costs by 5-10% and 25-40%, respectively.
Reduction in transportation and administration costs by 5-10% and 25-40%, respectively.
Customer reviews
,,
"Phytopharm Klęka S.A. is a pharmaceutical company that manufactures products, primarily based on ingredients of natural origin. We offer our customers finished products, as well as contract manufacturing services. Managing a diverse and rich assortment portfolio requires us to carefully analyze demand and continuously improve the planning process. Thanks to the artificial intelligence and machine learning mechanisms available in the Dature Premium application, we are able to diagnose demand changes faster and adjust inventory to market needs. As part of the Dature Premium implementation project implemented by Smartstock, functionalities for both tactical-strategic and operational horizon demand forecasting were launched. Vendor Manged Replenishment (VMR) inventory management optimization for selected suppliers has also been launched. This makes it possible to carry out integrated planning involving the management of Phytopharm products at our customers' distribution warehouses, forecasting demand there and optimizing the inventory of our own products. This generates not only a business benefit for us and our customers, but most importantly for patients, for whom we want to provide the best possible access to our products."
