Filter results

48078 results

Improving biotech production with real-time testing

Marieke Klijn detects changes in biotechnological processes in real-time, with the use of new monitoring techniques. She wants to move biotechnological research away from old-fashioned manual sampling to much more efficient continuous testing: “We can now know at once how many cells are alive during the production process, or how fast they are eating glucose, for example – all without the need to take a sample.” “My research focusses on ways to optimize production processes for biotech companies, by developing methods for continuous testing,” Marieke Klijn, assistant professor at the biotechnology department of the TU Delft, begins. “My team aims to make measurements an automatic part of production processes in biotechnology. I want to integrate sampling directly into the process – real-time testing: that way, we no longer need to rely on time-consuming manual handling in order to know how well the production process is going.” She set up her lab at TU Delft in September 2020. Read more Higher quality and more sustainable Klijn explains that continuous testing will lead to biotechnological products with higher quality and processing efficiency: “Real-time testing improves product quality, because the biotech company has continuous control of the product: the computer can easily detect deviations in the production chain at a much faster pace and change parameters so as to prevent failure, like a loss of product or unhappy cells. Real-time testing is also required if you want to move to continuous manufacturing, which allows companies to reduce their footprint and become more sustainable.” The biopharmaceutical industry is actively searching for more continuous processing setups that can lead to more efficient and affordable production of medicine, such as vaccines or cancer medicine. Klijn: “A continuous process flow already exists in other types of industries, such as the petrol industry. But in the case of biological matter, continuous processing and testing is more difficult: it has more technological, biological and regulatory challenges. You can’t be sure of the result in advance.” Eating habits of cells in real-time “In my lab we have a setup to insert analytical techniques into a bioreactor and extract processing data in real-time,” Klijn explains. “A bioreactor is basically a vessel with nutrients and cells. These cells can produce a specific biotechnological product, for example a food additive or a drug compound, in a highly controlled environment. We would like to monitor different cells to find out how to make these models robust for industrial application: so that when products or cell lines change, the company doesn’t need to go through the whole development phase again.” The main analytical tool that Klijn’s lab uses is Raman spectroscopy: a technique in which laser light is scattered due to molecular vibrations. Each molecule will have a different scattering pattern, making it possible to study all kinds of changes in different molecules in real-time: “With this technique we now know at once how many cells are alive during the production process, or how fast they are eating glucose, for example – all without the need to take a sample. The combination of this analytical technique and machine learning makes it possible to look at many different parameters at the same time.” The main analytical tool that Marieke Klijn uses is Raman spectroscopy, a technique in which laser light is scattered due to molecular vibrations. Blurry lines The biotechnological industry already makes use of Raman spectroscopy: they show a lot of interest in the models that the team can build for the process control systems. “We would like to capture as much information as feasible. For example, in addition to Raman spectroscopy, we want to use real-time imaging data to tell us how the cells are changing during the process of glucose eating.” The sheer amount of data this entails presents a challenge in itself, Klijn says: “Our dilemma here is how to effectively use the huge amount of information that we acquire.” How has this new focus on improvements during bio-production influenced the research field? “Before I started this work, each specialist would focus on their own part of the manufacturing process. Now the lines are blurry and they all work together. This makes the work very diverse. I can combine implementation of data analytical tools, with how the cell works and bioprocess engineering solutions. I work with experts from many different fields on the upstream part of processing, such as cell cultivation, and the downstream part such as modellers: all to make the production process into one single continuous flow.” Dr. ir. Marieke Klijn Assistant Professor +31 15 27 81280 m.e.klijn@tudelft.nl Room C0.550 Building 58 Van der Maasweg 9 2629 HZ Delft linkedin More stories

The Academic Fringe Festival - Aaron Halfaker: Designing to Learn - Aligning Design Thinking and Data Science to Build Intelligent Tools That Evolve

The Academic Fringe Festival - Aaron Halfaker: Designing to Learn - Aligning Design Thinking and Data Science to Build Intelligent Tools That Evolve 04 April 2022 17:00 till 18:00 - Location: Online by Aaron Halfaker | Microsoft Research Abstract “Design to learn" is a collaborative approach to developing intelligent systems that leverage the complementary capabilities of designers and data scientists. Data scientists develop algorithms that work despite the noisy, messy realities of human behavior patterns, and designers develop techniques that reduce noise by aligning interactions closely with how users think about their work. In this talk, I'll describe a set of shared concepts and processes that are intended to help designers and data scientists communicate effectively throughout the development process. This approach is being applied and refined within various product contexts in Microsoft including email triage, meeting recap, time management, and Q&A routing. Speaker Biography Aaron Halfaker is a principal applied research scientist working in the Office of Applied Research in Microsoft’s Experiences and Devices organization. He is also a Senior Scientist at the University of Minnesota. Dr. Halfaker’s research explores the intersection of productive information work and the application of advanced technologies (AI) to support productivity. In his systems building research, he’s worn many hats from full stack engineer, ethnographer, engineering manager, UX designer, community manager, and research scientist. He’s most notable for building an open infrastructure for machine learning in Wikipedia called ORES. His research and systems engineering have been features in the tech media including Wired, MIT Tech Review, BBC Technology, The Register, and Netzpolitik among others. Dr. Halfaker reviews and coordinates for top-tier journals in the social computing and human center-AI space including ACM CHI, ACM GROUP, ACM CSCW, Transactions on Social Computing, WWW, and JASIST. Homepage: https://www.microsoft.com/en-us/research/people/ahalfaker/ . More information In this second edition on the topic of "Responsible Use of Data", we take a multi-disciplinary view and explore further lessons learned from success stories and examples in which the irresponsible use of data can create and foster inequality and inequity, perpetuate bias and prejudice, or produce unlawful or unethical outcomes. Our aim is to discuss and draw certain guidelines to make the use of data a responsible practice. Join us To receive announcements of upcoming presentations and events organized by TAFF and get the Zoom link to join the presentations, join our mailing list . TAFF-WIS Delft Visit the website of The Academic Fringe Festival

Half Height Horizontal

Researchers hand over Position Paper to Tweede Kamer

On behalf of the TU Delft PowerWeb Institute, researchers Kenneth Brunninx and Simon Tindemans are handing over a Position Paper to the Dutch Parliament on 14 November 2024, with a possible solution to the major grid capacity problems that are increasingly cropping up in the Netherlands. The Netherlands is unlikely to meet the 2030 climate targets, and one of the reasons for this is that large industry cannot switch to electricity fast enough, partly because of increasingly frequent problems around grid capacity and grid congestion. In all likelihood, those problems will actually increase this decade before they can decrease, the researchers argue. The solution offered by the TU Delft PowerWeb Institute researchers is the ‘flexible backstop’. With a flexible backstop, the current capacity of the power grid can be used more efficiently without sacrificing safety or reliability. A flexible backstop is a safety mechanism that automatically and quickly reduces the amount of electricity that an electric unit can draw from the grid (an electric charging station or a heat pump) or deliver (a PV installation). It is a small device connected or built into an electrical unit, such as a charging station or heat pump, that ‘communicates’ with the distribution network operator. In case of extreme stress on the network, the network operator sends a signal to the device to limit the amount of power. Germany recently introduced a similar system with electric charging stations. The backstop would be activated only in periods of acute congestion problems and could help prevent the last resort measure, which is cutting off electricity to users. ‘Upgrading the electricity network remains essential, but in practice it will take years. So there is a need for short-term solutions that can be integrated into long-term planning. We, the members of the TU Delft PowerWeb Institute, call on the government, network operators and regulator to explore the flexible backstop as an additional grid security measure,’ they said. The entire Paper can be read here . Kenneth Brunninx Associate Professor at the Faculty of Engineering, Governance and Management, where he uses quantitative models to evaluate energy policy and market design with the aim of reducing CO2 emissions. Simon Tindemans is Associate Professor in the Intelligent Electrical Power Grids group at Faculty of Electrical Engineering, Mathematics and Computer Science. His research interests include uncertainty and risk management for power grids. TU Delft PowerWeb Institute is a community of researchers who are investigating how to make renewable energy systems reliable, future proof and accessible to everyone.

Empowering professionals – vital for the longevity of any organisation

Empowering professionals – vital for the longevity of any organisation “Employers need to recognise that allowing employees to develop themselves is essential to business success, and space must be made for that,” says Willem van Valkenburg, Executive Director of TU Delft’s Extension School for Continuing Education. In a recent Topic Talks interview on New Business Radio , van Valkenburg highlighted the need for a robust learning culture within organisations to keep pace with an evolving job market and rapid technological advancements. Barriers to continous development Reflecting on the learning culture in the Netherlands, as an example, van Valkenburg pointed out an often-overlooked barrier: although resources for professional development exist, business needs frequently take precedence, sidelining employee growth. “Employers must actively create environments that encourage continuous learning and foster dialogue about upskilling and growth opportunities. This is especially crucial for companies struggling with unfilled vacancies. When staffing is low, the demand on existing employees intensifies, making it harder to prioritise time for learning.” Recognising these challenges, TU Delft has developed short-duration online courses to help professionals fit learning around their work responsibilities. The importance of up-to-date skills is clear: businesses need to adopt new technologies to remain competitive, yet they often face skills gaps that traditional training does not cover. Van Valkenburg shared an example of a professional who, after completing TU Delft’s AI in Manufacturing course, applied their learning to increase production by 50%. To address the need for specialised knowledge, TU Delft’s approach goes beyond standard coursework, fostering innovation through collaborative learning communities. “Our learning communities bring together researchers, professionals, and policymakers to collaboratively address real-world problems. This structure enables participants to learn while actively solving practical challenges,” explained van Valkenburg. Throughout the conversation, van Valkenburg emphasised the value of aligning academic expertise with industry needs. By integrating TU Delft’s research into their courses, participants have access to the latest advancements, keeping them at the forefront of their fields. “Our courses, accessible in English and designed for a global audience, allow professionals from diverse sectors to stay ahead of technological shifts,” he said. Tangible impact In addition to cutting-edge content, the Extension School maintains a learner-centred approach. Feedback is integral, helping TU Delft ensure its courses remain relevant and impactful. “What moves me the most are people in developing countries who say, ‘I took a course on solar energy. With the knowledge I gained, I wrote a project plan to install solar panels in our village. Now, we have more electricity and can develop ourselves further.'" Building lifelong learning at all levels is essential to creating a resilient workforce and a sustainable future. TU Delft’s Extension School is committed to keeping these conversations alive, empowering professionals worldwide to drive meaningful change. For those interested in hearing the full interview in Dutch, please click the link. Luister hier Support Willem van Valkenburg and our work by giving us a shout-out on LinkedIn linkedin