The technology of machine learning (ML), a subset of artificial intelligence (AI) that allows experiential learning and improvement without being specifically programmed, is already providing significant benefits to the supply chain. If this trend of machine learning continues its current trajectory in manufacturing, we could eventually see complete end-to-end automation.
But the real benefit is in the supply chain management (SCM) area. This is an area where the same supply issues happen daily, they just appear differently. For example, delays are fundamentally similar, but where and when they occur is almost random. In addition, the data is now there to predict these, if only we had the human capacity to process that data in real time. The major benefit of ML in the supply chain is the ability to pair it with humans to learn from the human operation and scale so that for the next quality event, less human intervention is required and greater innovation can occur. In other words, human workers on the shop floor drive the agility in SCM to aid automation at all phases within the supply chain.
One way that ML is having an impact on demand planning is through its ability to identify certain patterns in datasets and detect anomalies for infectious disease intelligence.
The pharmaceutical industry has even more to gain through greater understanding of global weather conditions. ML applied to historical data can identify correlations between specific changes in climate and occurrences of a disease. If there is a wet summer forecast for a hot and humid area, for example, there is likely to be an increase in the mosquito population and therefore a higher likelihood of diseases related to mosquito bites.
In September 2017, researchers from the Saw Swee Hock School of Public Health and the National Environment Agency’s (NEA) Environmental Health Institute in Singapore developed an algorithm to forecast dengue outbreaks. They discovered that transmission dynamics of dengue, a mosquito-borne virus that infects approximately 400 million individuals annually, closely correlated with climate variables such as precipitation and temperature. By combining historic climate information with previous seasonal patterns of dengue, the NEA was able to predict outbreaks as early as four months in advance.
In addition, data provided by internet of thing (IoT)-enabled devices is helping to better monitor temperature controls and track products as they move through the chain. Merck, the global biopharmaceutical company, is among the most advanced in utilizing ML to optimize demand forecasting.
IoT can be used to aggregate data of batches from various production machines and systems to analyze and predict failure pattern. It helps a manufacturer ‘s supply chain function much like that of la self-driving car. A smart, automated manufacturing system can analyze data continuously and make decisions on its own about speed and resources. In manufacturing supply setting, a such a system would predict spikes and lulls in demand for products and suggest ways to reroute raw materials accordingly.
Veripad is an example of an organization that is using machine learning to solve issues surrounding the detection of counterfeit drugs in the supply chain. They provide a chemical test card capable of identifying components of common medications in minutes with no lab equipment. You just crush one tablet of the medication you want to test, rub it over the card, and soak the card in water. Veripad uses chemical testing paper and a mobile application to identify falsified medications.
However, they have developed their own testing cards that can perform several chemical tests on a medication in question by simply adding water. From there, they utilize their smartphone application to interpret the results of the chemical test card and indicate whether the medication is suspicious or authentic. Finally, they utilize data analytics technology to aggregate the results of every test in order to better understand the flow of counterfeits within affected regions with the ultimate goal to improve initiatives and policies.
#1 Define what you are using the technology for.
Before you even start building any ML model, you have to define upfront how it’s going to be used and why that’s going to make the experience better for a user.
#2 Make sure your data is ready.
Ensuring your data ecosystem is mature enough is critical. ML requires a strong pipeline of clean data. The difficulties with the data are many. Most companies undertaking ML projects already own and store vast quantities of data, but this data is often siloed and requires aggregating, which is a lengthy and difficult process that few have resources for. This data also needs to be secure, and there are many regulations in place that hold organizations back from using personal or sensitive data to train their ML algorithms, particularly in finance and health care. Hire the right people.
This is not an overnight process. It requires a significant investment, in terms of both finances and resources. According to a survey from Tech Pro Research, only 28% of companies have some experience with AI or ML, and more than 40% said their enterprise IT personnel don’t have the skills required to implement and support AI and ML.
#3 Be prepared for failure.
ML is still a nascent technology. It requires a significant investment in terms of finances and resources. Pharma supply chains must be ready for failure. Equally, this is also not a sector where failure is an option. This means testing is key.
#4 Collaborate with your supply partners.
Patient safety and drug efficacy must be the top priority when designing a supply chain strategy for pharmaceutical drugs – and manufacturers, distributors and providers all play important roles in delivering against this mission. Collaboration is vital for the success of ML – not just between supply chain partners, but pharma companies must also pool resources and ideas to ensure the technology reaches its full potential.
In summary, new digital design, manufacturing and Supply chain technologies offer pharmaceutical manufacturers the possibility of a systems-based, digital approach to substantially streamline their supply chain processes by adoption of the digital aspect of ML, AI and intelligent automation.
The aim is to integrate a wide range of predictive models to process data as key assets in the journey toward truly digital operations. The adoption of ML technology can systematically harness the potential of advanced materials, process modelling and data analytics to transform decision-making in product and supply chain process design and control.
Enjoying this blog? Learn More.
Pharmaceutical CGMP for 21st CenturyDownload Now