1st ACE Call Awards

Toward better future: Merging the lines between Fashion, Materials Science, Computer Science and Engineering

PI: Galina Mihaleva (ADM)
Co-PI: Ong Yew Soon (SCSE); Chen Xiaodong (MSE)


Due to increased reliance on creativity across industries as well as to the vaguely delineated boundaries of the concept of creativity, new trends have emerged as a way for interdisciplinary research efforts to innovate. One such complex new trend which have received more attention in academia is interdisciplinary creative work in fields such as fashion, science and technology. The growing popularity of wearable technology in fashion is proof of the effective symbiotic nature of seemingly incompatible fields of work. In recent years, terms such as creativity and creative work have largely become buzz words in the business context and are often used to further various arguments about the positive effects of the sector not only on the economy but also on our society as a whole. A significant number of scholars and other professionals from the cultural field have emphasized the potential of the creative industries in improving the overall quality of our life by encouraging people to express themselves and to share this experience with other like-minded groups (Bakhshi, McVittie, & Simmie, 2008). Essentially, the proposed interdisciplinary research idea in the realm of art, design and wearable experiences at the intersection of these tree disciplines: fashion engineering and material science among the School of Art, Design and Media, the Data Science and Artificial Intelligence Center (DSAIR) and the School of Materials research Science and Engineering aims to discover new solutions and new applications on smart data monitoring systems with novel materials that sticks to the body even while exercising, coupled with a cyber-system composed of wireless wearable sensors, data cloud and artificial intelligence capabilities of a highly conformal CEI that mechanically interlocks polymerization of pyrrole on RH-proof silk fibroin (SF). This current work not only endows CEI with highly conformal property, but our choice of SF also provides a stable adhesion at high RH. Last but not least, we believe that our proposed CEI serves to open up new opportunities and challenges of next- generation e-heathy monitoring, human-machine interfacial, and soft exoskeleton systems. 

Design and development of digital coaching environment for Deeper Experiential Engagement Projects: co-evolution of education with Artificial Intelligence

PI: Sze Chun Chau (SBS)
Co-PI: Wilson Goh Wen Bin (SBS); Tan Ooi Kiang (EEE); Christopher John Hill (SOH); Chua Bee Leng (NIE); Stefanie Chye Yen Leng (NIE)


Rather than writing off this “human” role of the advisor as irreplaceable by technology, we propose to develop mechanisms for AI to aid in increasing the number of students that can be coached per advisor while maintaining the advisor’s critical value-add. Essentially, we advocate not only making a shift in education focus because of AI as a disruptor, but working towards co-evolution of education with AI as an enabler. We propose to design and develop training sets and features that can ultimately be used to assemble an AI-based digital environment for “preliminary coaching”, in which a student’s journal submissions may be processed to provide advisors and students with profiles of the “lower order” learning gaps of students. This will free up time for advisors to channel the attention and efforts towards higher order formative feedbacks for communicating during the face-to-face interaction. Parameterization of inputs and development of training sets relevant to science, engineering, and humanities are required as DEEP group students from different Schools and cover themes that straddle various disciplines. These are further overlaid with feature designs anchored in pedagogy principles. As the study is driven by design thinking, it will particularly benefit from ACE funding as opposed to funding schemes that support more hypothesis-driven research. 


Tracking seeds and their animal dispersers in tropical forests: using engineering advances to address long-standing challenges in biodiversity conservation

PI: Norman Lim T-Lon (NIE)
Co-PI: Lee Yee Hui (EEE); Shawn Lum Kaihekulani Yamauchi (ASE)


Seed production and effective dispersal is crucial for plant reproduction and forest resilience. Animals are the dispersal agents for most tropical tree species, yet fruit-eating animals are at higher risks of extinction. There is little information on the success and extent of dispersal because of technical limitations to tracking seed dispersal and their animal dispersers at a scale that is appropriate for understanding forest-wide regeneration. This severely hampers our understanding and management of these threatened natural habitats for sustainability. Therefore, we propose to develop an automated wireless sensor network to track seeds and their animal dispersers in tropical forests. This proposed research is the culmination of a collaboration between an electrical/electronics engineer, a botanist, and a zoologist, breaking down traditional barriers between disciplines to solve this technical and intellectual challenge. Such interdisciplinary research to solve a real-world problem is perfectly aligned with the motivations behind the ACE programme. NTU is not only an ideal place to conceive and implement the proposed project, it may be one of the only institutions in the world with the necessary skillsets, experience, and institutional ties to carry this off. The project is in line with NTU’s Sustainable Earth Peak (the “peak of peaks” amongst NTU’s Five Peaks of Excellence), and falls under the Pillar of Technology Innovation under this Peak. By supporting this research, NTU demonstrates her clear commitment to research towards a sustainable and resilient Earth. This project is not suited for traditional competitive funding schemes because the funding is to support this proof of concept to effectively track large number of small items. While the proposal focuses on seed dispersal, the system developed can be readily applied in other areas (e.g., tracking disease vectors for epidemiology, and tracking of goods in logistic industry), potentially qualifying it as blue sky research.

Detecting and Tracking Volcanic Ash Using Social Media Data

PI: Cong Gao (SCSE)
Co-PI: Benoit Taisne (ASE)


This project will involve interdisciplinary efforts of Computer Science and Volcanology. It aims to solve the problem of detecting and tracking volcanic ash using machine learning techniques. Volcanic ash consists of fragments of pulverized rock, minerals and volcanic glass, created during volcanic eruptions. Combining the facts that volcanoes are present around the globe, with complex weather patterns, ash can have global impact on society, including human and animal health, disruption to aviation, disruption to critical infrastructure (e.g., electric power supply systems), primary industries (e.g., agriculture), buildings and structures.

In this project, we aim to investigate the problem of detecting and tracking Volcanic ash using social media data, such as tweets. Specifically, we have the following subobjectives: 1) researching and designing ways to expedite the processing of large data streams, performing real-time geoprocessing of incoming data; 2) linking external artefacts (media, internet websites) to each social media item, e.g., tweet; 3) integration with traditional source for volcanic activity to supports our data analysis and validate our efforts.To the best of our knowledge, no previous work has been done on detecting and tracking volcanic ash using social media data in real-time. For the first time, we will explore and invent new techniques based on machine learning combined with our knowledge about volcanic activity. We expect that the proposed solution in the project will significantly improve our overall ability to monitor and forecast the distribution of volcanic ash. The project should be funded through ACE programme, but not other traditional competitive funding schemes due to its interdisciplinary nature.


A Tale of Two Deficits: Causality & Care in Medical AI

PI: Melvin Chen (SOH)
Co-PI: Chew Lock Yue (SPMS)


Given certain demographic pressures and the imminent shortfall of healthcare resources, many have mooted medical AI as an alternative mode of healthcare. Two deficits must however be addressed: the causality deficit and the care deficit. Is it possible to operationalize Pearl’s (2000) Structural Causal Model, Mahadevan’s (2018) Riemannian manifold of imagined datasets, and related models of reasoning in terms of counterfactuals in AI systems? In addition, the best AI systems in the market lack the ability to care for patients that humans typically possess. Could approaches such as Van Wynsberghe’s (2013) care-centered value sensitive design approach help us to address the care deficit? These are the concerns that broadly circumscribe our project.

We are applying specifically for ACE funding rather than other traditional funding schemes, as our research ideas are bold, creative, and apply to the cutting-edge research domain of medical AI. State-of-the-art, machine learning-based AI is still committed to using statistical correlations from real-world datasets as a guide when making causal inferences. This keeps AI systems at the first rung of the three-level causal hierarchy: association (Pearl, 2018). We hope to break new ground by considering the prospects of designing AI systems that are capable of rising to the second and third rungs (viz. performing interventions and reasoning in terms of counterfactuals) of this hierarchy. Our initial aim is to focus on the causality deficit, but our longer-term aim (after the first two years of research) is to attend to the care deficit as well. Our project involves a unique collaboration between CoHASS (Philosophy) and CoS (Physics), when the general tendency has been to speak in terms of a chasm between the humanities and the sciences (Snow, 1959). Our hope is that an interdisciplinary appraisal of the causality and care deficits in medical AI will advance the field of medical AI research.