AI Advanced Analytics and IoT Sensor Technologies
Three examples of how the usage of artificial intelligence, advanced analytics, and IoT sensor technologies are being used in pharmaceuticals manufacturing to improve the supply chain and quality challenges are:
- Pfizer's collaborations with both IBM Watson Health and Concerto Health, to develop a cure for cancer.
- Bayer and Merck Co.'s combined AI software being used to identify and zero in on patients of CTEPH from a large pool of patients with similar results.
- AbbVie and AiCure's collaboration to test drug adherence, that resulted in 90% success.
Challenges in pharmaceutical supply chain management.
1. Drug Development time and cost:
- Drug development takes around 12 years to be complete.
- Only 10% of drugs are approved by the FDA.
- Developing a single new drugs costs around $2.6 billion. (Source3)
- Reducing the investment will result in slower development of drugs, thus slowing down the chain.
- To reduce the time, more funding as well as mental faculties will be necessary. This places severe restrictions on faster drug developments.
- AI can assess the large existing and oncoming pool of clinical and genomic data to find the best molecular combination for any specific disease.
- AI and machine learning, if introduced, will cut down the average production time by months and also reduce the cost of drug development, which for the 10 biggest pharmaceutical firms is close to $70 billion annually.
2. Drug Trials:
- Drug trials take a significant time to analyze and improvise their effects.
- They also usually have dangerous side effects.
- AI and machine will help select the best suited patient population for the trials, based on high volume data.
- By using robotic arms and automated testing, drug trials can be successfully accomplished on a much better rate.
- By using AI to assist the workforce, around $160 million could be saved on clinical drug trials.
3. Insights into rare diseases:
- Rare diseases have a very small pool of patients.
- Extracting data and trying to analyze patters can be very challenging.
- AI and machine learning tools can assess all the available data set and find patterns to better analyze a rare disease and develop a cure.
4. Quality Assurance:
- Quality assurance is another problem for pharma supply chain.
- Unprecedented failures and slowdowns also affect the quality of drug produced.
- Temperature sensitive drugs are often affected by this. (Source2)
- Implementing AI to track possible defects in manufacture as well predict failures and slowdowns can greatly speed up the manufacturing process as well as improve quality assurance.
- Using intelligent sensors to track storage temperature and adjust the temperature whenever necessary can also prevent drug waste.
Examples of Usage of AI, machine learning, advanced analytics in pharmaceutical manufacturing
1. Pfizer, cancer and AI:
- Pfizer is a research-based global biopharmaceutical company.
- Pfizer made a collaboration with IBM to use the Watson Health AI to further its oncology research.
- According to the Pfizer President of Innovative Health, Albert Bourla, "the best way to get the body to fight a tumor is some combination of agents to spur the immune system into action. But possible combinations are countless, so the greatest challenge is to find ways to narrow the field and predict what combinations might be more effective. Thanks to Watson, which has been trained to analyze historical data, we’re trying to predict the winning combination."
- IBM's Watson Health AI has already analyzed a lot of research data, which Pfizer can use to make non-obvious connections that could lead to combination medicines for cancer.
- Pfizer also entered into a collaboration with Concerto Health AI a month ago.
- These research collaborations will help Pfizer develop a cure for cancer faster.
- Since the collaborations and the products are still in development, results from both partnerships have not been made available.
2. Bayer and Merck AI Partnership:
- Bayer & Merck, are two of the largest pharmaceutical companies in the world.
- Bayer and Merck have developed an AI tool to collect image data from patients' pulmonary vessels, lung perfusion, and cardiac check-ups. The tool is able to identify CTEPH patients using the available images.
- This software was recently approved by the FDA.
- CTEPH is a rare disease caused by blood clots that don't dissolve in the lungs. Its prevalence is 5 per million people worldwide.
- CTEPH has similar symptoms to COPD and asthma. This makes it difficult for doctors to diagnose the disease.
- The AI tool uses data sets to differentiate it from diseases with similar symptoms.
- The AI software was approved a year ago and since the tool is still in development, no major breakthroughs or results have been shared with the scientific community.
3. Abbvie Collaborates with AiCure in using AI for Drug Adherence: (Source4, 10)
- Abbvie is a major pharmaceutical company. It recently partnered with AiCure, a company that uses AI and machine learning for pharmaceutical analysis.
- The goal of their collaboration was to analyze drug adherence.
- Abbvie used AiCure's facial and image recognition algorithm to monitor the adherence.
- This method resulted in an improvement of the adherence by over 90%.
- Drug adherence is very important for pharmaceutical companies to ensure that drug tests went according to set standards.
- The results of this collaboration prove that using AI can improve the trial's efficiency and correct procedural flaws and errors.
The interest in the use of AI, machine learning, advanced analysis, and IoT to enhance and better drug manufacturing and supply chain has increased in the last few years but is still a fairly recent phenomena. This creates limitations on the analysis of impact made by these new technological augmentations. These are due to three main reasons:
- Long development time — Despite AI reducing the time taken in drug development, the calculated time still takes around 10 years.
- High failure rate — The introduction of AI only reduces the error, but it doesn't eliminate it.
- Trade secrets — Due to competition among companies, rival companies may not present the exact data and results in order to stay competitive in the market.