|
Artificial intelligence (AI) is a crucial component that needs to be taken into account whenever we talk about Industry 4.0. The growth of computing and technology has permeated many aspects of science. Numerous facets of science have been impacted by the development of computing and technology. In all branches of science and technology, from fundamental engineering to pharmaceuticals, artificial intelligence (AI) is a crucial component of computer science. Growing investments in AI in medicine are fundamentally driven by the realization that data can and should be used more intelligently to make better decisions. Pharmaceutical companies may learn from prior data in ways that people cannot, and then use those learnings to optimize a range of processes using machine learning (ML).
As we have already shown, the industrial revolution is actually happening. Although every revolution has pros and cons, if we are positive, it will be beneficial. It is up to us to make the most of the advanced capabilities. Many individuals have the question in their minds. Is AI truly beneficial? Does AI eliminate jobs? We've returned to see the unfolding revolution. Just few decades back, people used to think that computers would take over occupations, but in actuality, they only widened the unknown horizon. If we recall correctly, in 1986 at Washington about 6,000 math teachers were passionately protested the introduction of calculators.
It is important to remember that the best can be used and that misuse should be avoided. Despite the fact that technology misuse is widespread, it is always present in any technical breakthrough.
In fact, traditional methods have a role to play in development and production, and they can be used to strengthen AI's support for human intelligence, creativity, and invention. Meanwhile, enormous effort and a significant increase in IT infrastructure is required. AI has the ability to reduce operational cycle times while increasing quality and/or lowering total expenses and raw material consumption.
The first industrial revolution (Industry 1.0) was powered by steam; the second industrial revolution (Industry 2.0) was powered by electricity; the third industrial revolution (Industry 3.0) was powered by early automation and machinery; and the fourth industrial revolution is being shaped by intelligent computers (Industry 4.0).
Industry 4.0, the fourth industrial transformation, pools AI, ML and big data to allow integrated and automated production systems. As we predict Industry 4.0 is to be a massive transformation. The increasing complexity of developing novel medicines with demand of shorter launch time to market, fuel the need for faster research and manufacturing. Such a present scenario demands the incorporation of AI into a wide range of pharmaceutical research and manufacturing operations.
Artificial intelligence refers to systems, computers, and/or technology, such as algorithms, that "mimic human intellect to complete tasks." AI is also referred to as a self-learning technology, meaning equipment and/or technologies that "can interactively improve themselves, based on the information they acquire". This is known as ‘machine learning' or 'deep learning' and it allows smart machines to function (example: deep Q-learning robotics).
The AI being used in pharma and other industries is a narrowly focused type of machine intelligence designed to solve a specific task or set of tasks using automated algorithms.
AI can aid in the improvement of processes that require a large number of people, ranging from maintenance experts to quality control and beyond. AI tools can improve output by automating the most complex functions. It guarantees that duties are completed precisely. In addition to producing high-quality work, it can analyze processes, identify weak points, enhance decision making, and identify areas that can be simplified.
However, new technologies are usually costly to employ in their early stages of development/implementation. They also carry with them unknown dangers. This may make pharmaceutical companies reluctant to embrace AI. Despite these obstacles, the potential savings associated with AI, such as reduced human mistakes and increased detectability of deviations and OOS findings, make it an appealing move into the future of diagnostic medical care, drug development, and treatment choices.
AI and machine learning will continue to aid in medication finding and production. And, as AI tools become more available over time, they will become a natural component of the pharmaceutical and industrial processes. AI will be used in the future. Given current patterns and the rising role of AI in pharma, businesses may need to consider new regulatory boundaries when undertaking research using AI technology.
According to a recent Deloitte research, 75% of major firms (> $10 billion in annual sales) invested more than $50 million on AI projects/technologies. Even smaller firms (annual revenues ranging from $5 to $10 billion), 95% of such mid-sized organizations invested up to $50 million by 2020.
The worldwide AI software industry is expected to rise from $10.1 billion in 2018 to $126 billion by 2025, according to Tractica.
According to McKinsey Global Institute study, the impact of AI and ML on the pharma market was expected to produce approximately $100 billion across the US healthcare system by 2021. According to the McKinsey Global Institute, AI and ML in the pharmaceutical business could produce nearly $100 billion per year across the US healthcare system.
Global AI earnings in the pharmaceutical, medical, and healthcare industries are projected to hit nearly $21 billion by 2025, according to GlobalData. Pharmaceutical Technology's main firm is GlobalData.
Global Data, which employs an AI deal and patent tracker in the pharmaceutical sector, recently published some intriguing observations involving AI or comparable technology transactions in the industry over the previous nine quarters. According to the study, while AI has yet to reach its maximum potential, it will be the most disruptive technology in the drug industry by 2022. Since October 2021, there has also been a substantial rise in hiring for AI jobs in this industry. According to current market data, the adoption of AI in the drug discovery industry is expected to grow at a CAGR of 36.1% between 2021 and 2031. Cybersecurity was the top tech expenditure in the pharmaceutical business in 2021, followed closely by e-commerce, big data, and AI.
As per ‘Research and markets’ report the worldwide AI in pharma market is expected to be worth roughly $699.3 million in 2020, with a compound annual growth rate (CAGR) of 31.8% since 2015. The market is projected to expand at a 32.9% annual rate from $699.3 million in 2020 to $2,895.5 million in 2025. The market is projected to expand at a CAGR of 25.9% beginning in 2025 and reaching $9,142.7 million by 2030.
AI in pharmaceutical technology transfer
AI may be used in the pharmaceutical manufacturing process to increase productivity, efficiency, and speed up the creation of life-saving medications. Every parts of the manufacturing process, including quality control, predictive maintenance, waste reduction, design optimization, and process automation, may be managed and improved using AI. AI can replace time-consuming traditional production procedures, allowing pharmaceutical companies to bring pharmaceuticals to market faster and at a lower cost. Apart from significantly enhancing their Return on Investment by eliminating human interference in the production process, AI would also remove any possibility of human mistake.
Lonza, a Swiss CDMO, gathers and analyses production data in order to optimize operations, develop process insights, boost yields, and satisfy quality standards. Lonza discusses how AI, ML, and big data are improving safety, quality, and sustainability while lowering costs. Lonza is investigating the potential applications of AI and ML in product technology transfer. Lonza faces a variety of sizes and equipment configurations during technology transfers. The number of process variables and important quality requirements involved in technology transfers adds another layer of complexity. AI and ML solutions are intended to estimate process performance or crucial process phases in such technology transfers, supporting in tackling these complex issues.
AI offers several chances to optimize processes in development and manufacturing. AI can do quality control, decrease material waste, increase production reuse, and perform predictive maintenance, among other things.
AI may be applied in a variety of ways to increase industrial efficiency, resulting in faster output and reduced waste. A procedure that generally requires human interaction to enter or maintain process data, for example, can be done using CNC (computer numerical control). The AI ML algorithms not only guarantee that tasks are completed exactly, but they also assess the process to identify areas where it may be simplified. This leads in less material waste, faster manufacturing, and more consistent meeting of the Key Quality Attributes of the product (CQAs).
AI can be designed to perform a variety of tasks; it is only restricted by nature and the constraints of its programming. The fundamentals of drug manufacturing in the pharmaceutical industry mandate that the company must create drugs in large amounts in order to be lucrative and recoup research costs. This, however, necessitates speed and scalability at every stage of the production process. Most of these problems can be addressed by incorporating AI and ML into pharmaceutical production. AI and ML have the potential to improve productivity at every stage of the process, from research and development to production and distribution. It can prevent scarcity by using predictive AI to anticipate demand. It is even possible to optimize supply networks by combining real-time view of cargo position, pace, and movement with traffic and weather data to forecast the best path and an exact ETA for storage.
Several pharmaceutical companies are utilizing Industry 4.0 solutions such as robotics and sophisticated analytics. SmartX automated process management technology, for example. SmartX analyses massive amounts of data collected during the manufacturing process and uses advanced data analytics and ML to generate crucial insights. This allows producers to spot process bottlenecks and remove needless obstacles, resulting in cheaper and more optimized output. Because numerous systems in the pharmaceutical industry generate massive amounts of data, it can be difficult for producers to remain on top of their process verification reporting. AL and ML-based solutions include contextualization and deep learning skills to help with data integration, normalization, and analysis, allowing a Continuous Process Verification (CPV) environment. The transfer of data and expertise from various sources is assisting producers in improving the batch drug-making process through resource optimization. AI provides the best drug production configuration with minimum waste, from predicting the various resources required for a drug to projecting the best timetable for producing it. AI can detect flaws in raw materials even before they reach the manufacturing process, ensuring quality and legal conformance. Furthermore, the Internet of Things (IoT) combined with AI technology is assisting makers in detecting faulty goods, packaging, and equipment.
AI-based computer vision systems have a variety of consequences in pharmaceutical production. When manufacturing drugs, physically ensuring the quality of each drug product can be a time-consuming and tedious job. Image processing is used by computer vision-enabled systems to study pharmaceuticals on conveyor lines in order to identify flaws (differences in form and color) at a much faster and more accurate rate. They can also identify flaws in the medication packaging. This also enables pharmaceutical firms to eliminate potential contaminations produced by human touch.
Other applications of AI
AI has the potential to be a game changer in this field by assisting producers with the initial screening of drug molecules in order to forecast the success rate of formulas. According to a market study, AI in drug development will be worth $1,434 million by 2024. Cross-industry alliances and ecosystem relationships, as well as the need to reduce medication transportation time and costs, would be the primary drivers of this development. The rising demand for control in the drug development process, as well as the increased usage of Cloud-based apps and the nearing patent expiration of some of the world's most critical medicines, would help fuel this rise. Deep learning capabilities of AI and the processing of vast volumes of data are significant driving forces that make AI a huge winner in drug discovery. Researchers are using algorithmic pairings, computer visualizations, and neural networks to evaluate microscopic pictures, build novel chemical representations, and even identify new paths for organic synthesis in the literature.
With its Machine Learning for Pharmaceutical Discovery and Synthesis consortium, Massachusetts Institute of Technology (MIT) collaborated with Novartis and Pfizer in 2018 to change the drug design and manufacturing process.
Commercial models of AI
Rethinking the commercial model is clearly a major goal for today's life sciences executives, with many already increasing their investments in AI to fulfil the demand for contemporary healthcare systems and client online engagements. While there are several AI-powered use cases to consider, it is critical for life sciences firms to choose those that provide immediate value while also promising long-term, sustainable, and scalable benefits.
The Werum PAS-X MES (Manufacturing Execution System) Suite digitally controls, monitors, and documents operations throughout the production cycle. It facilitates the reduction of mistake rates and production costs. Moreover, it reduces time to market and increases efficiency. Werum PAS-X MES covers the entire pharmaceutical, biotech, and cell and gene therapy production cycle, from process development to commercial production and packaging. The production management system incorporates all critical manufacturing tasks in a modular fashion. PAS-X is ready to use and seamlessly connects all surrounding IT systems via standard interfaces. i The Werum PAS-X Intelligence Suite makes data-driven decisions to optimize the manufacturing process. Werum PAS-X Intelligence Suite collects and contextualizes key production data from many sources, making it available for easy and rapid usage. A variety of technologies may be used to analyze and evaluate this data, resulting in a clear display of the results. If any difficulties or delays arise, clever algorithms will alert you in a timely manner, allowing you to take suitable countermeasures. This allows the client to increase yield, lower manufacturing costs, and make better use of active components. The AI and predictive analytics software Werum PAS-X Chromatography Control from Körber Pharma enhanced process performance. PAS-X Chromatography Control's AI is based on Deep Learning, a type of machine learning that uses an artificial neural network design. Using past chromatography data as a reference, data science and AI professionals first train the neural network. The data-adapted prediction model is subsequently evaluated, tested for predictive quality and other quality requirements, and integrated into the purification process. PAS-X Chromatography Control, when combined with a defined set of rules and current chromatography data, may forecast the best time for product fractionation based on projected product yield and purity with a prediction quality of more than 95%.
Regulatory perspective of AI
Along with significant AI-related deals in the pharmaceutical business, regulators tried to implement safeguards to avoid potential bad impacts of this technology.
The United Kingdom is attempting to establish worldwide AI standards. The Alan Turing Institute, the British Standards Institution, and the National Physical Laboratory established a collaboration in January 2022 to develop worldwide technical standards for AI. This plan and the new standards that result from it seek to better privacy standards for any data used by AI technology and reduce data biases. In pharmaceutical research, this could help protect the private of patient data used by AI systems while also reducing bias due to race, gender, and other variables.
The FDA and the Office of Digital Transformation released their Cybersecurity Modernization Action Plan in 2022, with the goal of "strengthening the FDA's ability to protect sensitive information, modernizing cybersecurity capabilities, and improving situational awareness to reduce overall security risks to the Agency." This plan aims, among other things, to improve the cybersecurity of AI systems.
In the European Union, the planned AI Act, billed as the first AI-specific legislation from a significant authority, divides the risk of various AI apps into three groups based on the degree of risk they pose. The European Commission adopted proposals in September 2022 to update and harmonies liability rules for manufacturers, including those in pharma, on a national level, so that they complement the AI act.
In November 2022, the Digital Charter Implementation Act (DCIA) was introduced in Canada's House of Commons, with one section dealing with AI. The declared goal of the AI and Data Act (AIDA), which is included in the DCIA, is to "regulate international and interprovincial trade and commerce in AI systems," as well as to suggest guidelines for the design, development, and use of those systems.
In March 2022, the Chinese government declared an obligatory registration system under which AI algorithms with "public opinion characteristics" and "social mobilization capabilities" must be submitted within the Internet Information Service Algorithm Filing System. China's algorithm regulation focuses primarily on the role recommendation algorithms play in information dissemination, ensuring that providers do not "endanger national security or the social public interest," and requiring them to "give an explanation" when they harm users' legitimate interests. In the pharma context, China's law may require businesses to submit pertinent algorithms, but most experts believe that this is primarily intended to prevent the ethical risks and biases of certain AI algorithms.
Conclusion
Leading pharmaceutical businesses are embracing AI technologies to remain ahead of the competition, such as AI-powered drug discovery platforms or manufacturing tools. Companies that are adopting AI include Johnson & Johnson, Bristol-Myers Squibb, AbbVie, AstraZeneca, Pfizer, Roche, Merck, GSK, and Sanofi. As AI becomes more firmly integrated in the healthcare business, an increasing number of organizations are using AI to avoid falling behind. The pharmaceutical business may easily cooperate on global drug research, drug discovery, knowledge transfer, scale-up, clinical trials, and production with the correct ML platform. The top challenges faced by organizations include high cost of AI systems (36%), AI integration into the organization (30%), AI implementation challenges (28%), data challenges (28%), selection of right AI technologies (27%), AI risk management (28%). Would AI aid in the transfer of pharmaceutical technology? It most certainly would. As previously noted, various pharma businesses are adopting new AI-based technologies for a variety of purposes, and with many more AI-based technologies still in development and on the horizon, the future of pharma will be AI-based as well. The only constant is change.
(Authors Hitesh Chavda & Ketan Savjani are with Emcure Pharmaceuticals Ltd, Mehsana, Gujarat, and Amol Deshpande & Atul Phatak are with Department of Pharmaceutics, PES Modern College of Pharmacy, Nigdi, Pune)
|