Events Calendar
-
[ May 07, 2019]
Green Electronics to Gray Matter
| While the computing demands of Information Technology are ever increasing, the capabilities of electronics have hit fundamental walls due to energy and dimensional unscalability. In this talk, I will demonstrate the quantum mechanical transistor, that I developed [Nature 2015], which beats the fundamental energy limitations. This device is the world's thinnest channel (6 atoms thick) sub-thermal tunnel-transistor. Thus, it has the potential to allow dimensional scalability to beyond Silicon scaling era and thereby to address the long-standing issue of simultaneous dimensional and power scalability.
Going beyond electronic computation, I will discuss about the biological computer: the brain, which can be thought of as an ultimate example of low power computational system. I will introduce the technology [Soc. Neuroscience 2016], which reveals for the first time, a nanoscale trans-synaptic architecture in brain and the way mother nature has engineered biomolecular organization in the brain to optimize its computing efficiency.
I will conclude with my research vision for how extremely powerful technologies can be built by fusing diverse fields and discuss briefly about the research directions of my new lab at MIT. -
[ May 06, 2019]
Precise and Reliably State Estimation for Connected and Autonomous Highway Vehicles
| Accurate and reliable awareness of world interactions is a key requirement for effective commercial deployment of autonomous and connected vehicles. Awareness arises from onboard sensors and ubiquitous communication between vehicles and infrastructure. Vehicle coordination and safety necessitate reliable “where-in-lane” knowledge of vehicle state. This presentation will address sensor fusion for high-bandwidth vehicle state estimation with a focus on high accuracy and reliability.
Advances is sensing and computation have dramatically altered the focus of related research. For example, computer vision and Global Navigation Satellite Systems each separately provide far more measurements than are necessary for observability. Such environments are signal-rich. The large number of measurements provides both opportunities (e.g., high accuracy) and challenges (e.g., large numbers of outliers). Standard state estimation approaches that decide irrevocably at each time which measurements are valid (e.g. EKF) are not sufficiently reliable at removing the effects of spurious measurements. When that decision is wrong, either measurement information is lost or the state and covariance estimates become corrupted, rendering all subsequent decisions suspect. Either situation can result in divergence of the state estimate, with potentially tragic consequences.
This presentation will consider moving horizon nonlinear state estimation by a novel risk-averse performance-specified (RAPS) approach. Moving horizon methods extract the Bayesian optimal trajectory using all sensor data over a temporal window (e.g. SLAM and RHE). RAPS modifies the optimization problem to select the least risky set of measurements that satisfies a user-defined performance constraint. RAPS is able to evaluate, and reconsider, outlier assumptions for all measurements within the temporal window. The presentation will include experimental results. -
[ April 19, 2019]
Journey from Energy Harvesting and 4D Printing to Medical Applications
| Energy can be harvested from vibration and human motion. Piezoelectric and electromagnetic power generators were used to transform the mechanical energy from vibration and human motion into electrical energy. On the other hand, robotic exoskeletons that can assist people with impaired mobility have been developed. With the developed device, paralyzed individual can regain the ability to stand up and walk. Smart ankle-foot prostheses with compact cam-spring mechanism have also been implemented to help amputees walk with less effort while having more natural gait. Utilizing additive manufacturing into smart materials has led to 4D printing technology for creating structures that can change their shape and function on-demand and over time. Actuator units were designed and fabricated directly by printing fibers of shape memory polymers in flexible structures. They can serve as tubular stents and grippers for biomedical applications. In this talk, related research projects and key results will be presented. -
[ April 09, 2019]
Energy Loss Processes in Organic Solar Cells
| Organic solar cells have the potential to revolutionise the way we harness the power of the sun. They are based on earth-abundant non-toxic raw materials and can be processed by inexpensive and scalable production technologies like roll-to-roll printing or vacuum processes on flexible substrates and on large areas. However, organic solar cells are still significantly less efficient compared to their inorganic counterparts. While organic molecules can absorb light much better than silicon, charge carrier extraction as well as energy losses at open circuit seem to limit the power conversion efficiency of organic solar cells. Over the past years, a particular focus has been put on a better understanding of the latter, highlighting the role of the charge transfer states in the generation and recombination of charge carriers as well as limiting the open circuit voltage, which will be the focus of this presentation. -
[ March 26, 2019]
A New Approach to Sequential Stopping for Stochastic Simulation
| Simulation is a powerful numerical tool set for performance evaluation and optimization of stochastic systems. Successful implementation of this approximation scheme requires one being able to assess the quality of the estimators and control the estimation errors. In this talk, I will present a new sequential stopping framework for stochastic simulation problems in which variance estimation is difficult. Examples include steady-state simulation, parameter estimation using stochastic gradient descent, quantile estimation etc. The proposed procedure provides statistical guarantees for the quality of the estimators. This is joint work with Peter Glynn. -
[ March 20, 2019]
RSC Journal Publishing Workshop
| Dr Sam Keltie will give an overview on his role and the scopes and review process for RSC journals, focusing on the materials & nanoscience journals. There will be time for Q&A throughout the workshop. The workshop is open to all members of the PKU community. -
[ March 08, 2019]
Similarity and Distance Between Networks
| In this talk, we discuss techniques for measuring the structural distance/similarity between networks. We start by outlining very basic concepts to do so and elaborate on some historical developments. Afterwards, we outline a novel technique for comparative graph analysis using topological indices and draw some conclusions. -
[ March 08, 2019]
Building Toward the Digital Twin for Smart Logistics Operation
| In this talk, we will introduce some new initiatives in Singapore in transforming the industry into the next generation industry, in particular in the logistics and maritime sectors. We will also present some of our latest work in developing a simulation optimization framework for digital twin for various systems, such as warehouse system and port system. This framework O2DES has the capability in incorporating optimization in the simulation, and optimizing the system parameters based on simulation output. -
[ March 08, 2019]
AGV Routing and Control for Transshipment Container Terminal
| In this presentation, we develop a framework for routing and controlling the Automated Guided Vehicles (AGVs). As the AGV routing and control are done in real time, the framework requires the solutions to be generated very fast. To address this computational challenge, the solution framework consists of two stages. The first stage determines the rough routes of the AGVs. The second stage determine the detail routings and schedules of the AGVs. As solving the detail routings and schedules require high computing resource, we decompose the port layout into submodules where each of them can be solved almost independently. The output of the first stage (rough routes) becomes an input to the second stage which solve the detail routes and schedules for real time control. Whenever an AGV completes a task, the first stage will be called to revise the rough routes of the AGVs which includes the new task which is assigned to the AGV that has completed the previous task. The new rough routes will become input for second stage to solve the detail routes and schedules. We show this framework is promising and is able to generate the detail routes and schedules very fast. -
[ Janurary 07, 2019]
Turbulent/non-turbulent interfaces and their relation to turbulent entrainment
| In many relevant flow cases, such as in jets or turbulent boundary layers, turbulent regions of the flow are bounded by non-turbulent fluid. The interface between such regions, the so-called turbulent/non- turbulent interface (TNTI) is known to remain sharp and well-defined even after long times. Here, we investigate how processes at the TNTI are related to turbulent entrainment, i.e. the incorporation of previously non-turbulent fluid into the turbulent flow. We show that turbulent entrainment is a viscous, small-scale process that is amplified by interface contortions across a range of scales. We also elucidate how these processes change in the presence of a stable stratification across the TNTI.