《clustermarket：2024自動化和人工智能在研究中的應用-當前應用和未來展望白皮書（英文版）（11頁）.pdf》由會員分享，可在線閱讀，更多相關《clustermarket：2024自動化和人工智能在研究中的應用-當前應用和未來展望白皮書（英文版）（11頁）.pdf（11頁珍藏版）》請在三個皮匠報告上搜索。

1、Automation and artificial intelligence in research-current applications and future perspectivesWhite paper|Automation and artificial intelligence in research-current applications and future perspectives|June White paper:Automation and artificial intelligence in researchAs technology continues to per

2、meate every aspect of our lives,terms like“automation”and“artificial intelligence”are becoming more prevalent,especially in the field of research.Despite often being used interchangeably,these two concepts are distinct and have different applications in laboratory settings.“Automation”relates to the

3、 use of technology to perform tasks that would typically require human input.1 Conversely,“artificial intelligence”(AI)is the use of intelligent systems to perform tasks requiring human intelligence,such as problem-solving and decision-making.1 Although distinct in concept,both have the power to rev

4、olutionize scientific ContentsIntroduction.4Current trends in automation.5 Data management.5 Sample management.5 Mobile robots and self-driving labs.6 Cell culture automation.6Current trends in artificial intelligence.7 Pathology.7 Drug discovery.7 Clinical trials.7 Gene therapies.8Current challenge

5、s and future IntroductionIn a previous edition of our panel discussion series,Lab Talks,we delved into how automation and artificial intelligence are accelerating scientific discoveries.Despite the discussion taking place over a year ago,the points raised still hold true today.Our panelists-Joana Ro

6、cha,a bioengineer&research fellow at INESC Technology and Science;Fane Mensah,life science business director at Synthace;and Kimberly Holtz,sales and business development manager at Iris.AI-all agreed that automation and AI have the potential to increase research efficiency and mitigate the risk of

7、human error.By taking over monotonous,time-consuming,and error-prone tasks,these tools give laboratories more time and resources to focus on tasks that require human expertise.This speeds up the discovery and innovation process.In this white paper,we review current trends and applications of automat

8、ion and AI in research laboratories,ranging from mobile robots and self-driving laboratories to“digital twins”and automated disease In any research project,data is crucial,but traditional methods of manual data collection and processing can be both time-consuming and error-prone.As laboratories face

9、 increasing demands for high throughput,there has been a surge in the development of automated data management processes.2 These methods enable large amounts of data to be collected,aggregated,validated,and translated with minimal human intervention,saving time and producing more accurate results.2O

10、ne significant development in the automation of data handling is the rise of Internet of Things(IoT)technologies.These wireless devices can be attached to equipment and collect both experimental data and metadata,providing context information to improve reproducibility and aid with troubleshooting.3

11、 Researchers can also remotely monitor experiments,allowing for multitasking while experiments are running.Current trends in automation Data managementWatch our panel discussion:Lab Talks Live While full AI utilization across a body of research can be a difficult and lengthy process,it could potenti

12、ally boost scientific discoveries across many fields.Sample managementJust like data,samples are crucial for any scientific experiment.Large laboratories often have millions of samples,which must be stored,retrieved,and processed efficiently and accurately.Unfortunately,this can create bottlenecks i

13、n laboratory workflows.Therefore,there is a growing need for automated sample management to help address this issue.4Implementing sample-specific 2D barcodes and a barcode capture system is a straightforward yet highly effective way to prevent costly and hazardous sample mix-ups.As technology contin

14、ues to progress,automating more complex sample management tasks,such as liquid sample aliquoting and DNA extraction,is becoming more feasible.4 Incorporating these solutions into sample management workflows offers significant benefits,including improved precision,safety,and laboratory efficiency.The

15、se advantages extend across virtually every field of Although stand-alone automated instruments have been used in labs for years,transporting samples and equipment between stages of processing still requires human intervention.However,as robotic technology advances,mobile robots are transforming lab

16、oratories into fully self-driving entities.5Mobile robots are devices designed to transport labware.They typically come with a base that moves around the lab,a robotic arm to carry cargo and various sensors to prevent collisions.By minimizing the need for human intervention,these robots increase pro

17、cessing efficiency and make it possible to run complex laboratory protocols overnight.5Mobile robots and self-driving labsCell culture automationstraightforward procedures,such as media replacement.7 This approach frees up human resources,allowing for more complex tasks to be carried out manually.Ce

18、ll culture is a highly versatile and essential technique used in scientific research to study the biology,physiology,and biochemistry of cells.Researchers can observe how cells respond to external factors in their natural environment,making it applicable to numerous scientific fields.6 Due to strict

19、 standardization and the increasing demand for efficiency,more cell culture laboratories are adopting automated workflows,which not only save time but also enhance sample quality.7Automation systems for cell culture labs differ depending on the labs needs.Some labs have fully automated the cell cult

20、ure process with robotic arms,while others use technology to automate only the most frequent AI is revolutionizing the field of pathology by providing the potential to automatically classify cells into different groups,and identify patterns and cell conditions to diagnose patients.This helps to stre

21、amline pathology processes,leading to faster,more accurate diagnoses and reducing the likelihood of human error.8Many international diagnostic companies are now utilizing AI to optimize their digital pathology procedures,and numerous start-ups have begun developing AI tools to detect various disease

22、s.8 One recent example of this is X-ZELL,which uses blood single-cell imaging combined with AI to detect early-stage cancers.A recent pilot study of their technology resulted in a 70%reduction in unnecessary biopsies.9Current trends in artificial intelligence:PathologyDrug discoveryOne of the most p

23、opular applications of AI in recent years has been in accelerating drug discovery by sifting through millions of drug targets and compounds to identify leads,as well as by providing quicker validation and optimization of drug target and drug structure design.10 This has led to a rapid increase in pa

24、rtnerships between AI-powered drug discovery companies and traditional pharmaceutical businesses.11For example,biopharmaceutical giant AstraZeneca has formed a multi-year collaboration with the clinical-stage AI-enabled drug discovery company BenevolentAI to create“living maps”of disease,which has r

25、esulted in the identification of multiple gene targets for idiopathic pulmonary fibrosis.12According to data collected from the Boston Consulting Group,“biotech companies using an AI-first approach to research had more than 150 small-molecule drugs in discovery.”Over 15 of these are already in clini

26、cal trials,and this pipeline has been seen to be expanding at almost 40%per year.40%Gene therapiesGene therapy has enormous potential to improve patients lives by eliminating the need for daily disease management.15 However,the creation of complex biological molecules,such as plasmids and sgRNA,to e

27、dit the genome is one of the most difficult aspects of the field.16 To tackle this challenge,AI technology has been developed to simplify the process and improve gene therapy workflows.Clinical trialsAs well as optimizing drug discovery workflows,AI is also being used in clinical trial design to imp

28、rove drug trial success.A popular use for AI in accelerating drug approval is in cohort selection,where AI is used to analyze vast amounts of epidemiological and historical clinical data to quickly identify individuals who may benefit most from a particular drug trial.13AI technology is also being d

29、eveloped to completely replace participants.For example,Unlearn.AI is using AI to develop“digital twins”of trial patients,predicting health outcomes based on different treatment options and thereby reducing control group size while maintaining data quality.14 Not only does this reduce costs for drug

30、 developers,but it also alleviates ethical concerns by limiting the number of placebo-receiving individuals.For example,researchers at the NYU Grossman School of Medicine and the University of Toronto have recently developed a universal model for zinc finger design,16 while the University of Illinoi

31、s has developed PlasmindMaker,an automated,high-throughput platform for plasmid design and construction.17 By speeding up this usually time-consuming and error-prone stage of the gene therapy workflow,these AI technologies Current challenges and future perspectivesThe use of automation and AI has ex

32、panded significantly over recent years,providing optimization for all stages of the research process,from experiment design to data analysis.This expansion has been driven in part by an increased shortage of workers and high demand for short processing times.18 As technology continues to advance,we

33、can expect to see this trend continue,with previously manual processes becoming automated,improving laboratory efficiency,accuracy and safety.Drug discovery,genomics and proteomics are research areas where we may see a significant increase in the use of automation and AI in the coming years,due to t

34、he high costs and risk of error associated with these disciplines.18 It is clear that automation and AI provide significant advantages for researchers.However,there are a number of challenges associated with the adoption of these technologies.One challenge is the initial high time and monetary costs

35、 associated with implementing these new technologies,resulting in a temporary reduction in efficiency.19 Additionally,when implementing technology into laboratory workflows,there is the risk of breakdowns and software issues which can lead to delays and potential sample loss.19 These caveats can mak

36、e researchers hesitant to implement automation and AI technologies in their laboratories.As there is no“one-size-fits-all”approach to laboratory automation,it is important for each laboratory to assess its own workflows and determine where it makes the most sense to implement automation and AI to le

37、verage their benefits with minimal risk.Read our blogon AI and automation While the exact definition and parameters of the lab of the future havent been determined,it is clear that with technology and the wider environment continuing to evolve and innovate at greater speeds,theres still plenty of sc

38、ope for improvement in References1.Difference between Artificial Intelligence and Automation-GeeksforGeeks2.What is Automated Data Processing?3.Scientific Data Collection for Laboratory Equipment:Which Method is Best?4.Automated Biobanking:Advantages,Benefits and Future Directions5.An intro to mobil

39、e robots and why they are useful in the lab|Biosero6.Cell Culture-PMC7.Cell Culture Automation:An Overview8.How Artificial Intelligence Can Revolutionize the Diagnosis Process in Pathology9.Application|X-ZELL10.Artificial intelligence in drug discovery and development-PMC11.AI in biopharma research:

40、A time to focus and scale|McKinsey12.Exploring genetic drivers of disease to transform the future of IPF 13.AI benefits in patient identification and clinical trial recruitment has challenges in sight14.Digital Twins 15.Gene Therapys Promise:Future Uses,Applications&Prospects|Pfizer16.A universal de

41、ep-learning model for zinc finger design enables transcription factor reprogramming|Nature Biotechnology 17.PlasmidMaker is a versatile,automated,and high throughput end-to-end platform for plasmid construction|Nature Communications 18.Laboratory Automation-2023-Wiley Analytical Science19.The Opportunities and Challenges behind Lab AutomationClustermarket Ltd1 Poultry LondonEC2R 8EJUnited K