Seminars: 2017 | 2016 | 2015 | 2014 | 2013 | 2012 | 2011 | 2010
Mechanism and Sensor Design for SUPERball, a Cable-Driven Tensegrity Robot. Drew (Andrew P.) Sabelhaus, Fri. Dec 15th 2:00pm, 230 Hesse Hall
Abstract: Mechanical design for cable-driven robots, especially those in tensegrity (tensile-integrity) tension networks, introduces a variety of challenges not found in other types of robotics; in particular, novel problems exist for cable routing, sensing, and actuation. This work describes designs for SUPERball, the Spherical Underactuated Planetary Exploration Robot Ball, constructed at NASA Ames Research Center. SUPERball is a proof-of-concept spherical tensegrity robot built to show dynamic rolling locomotion. Engineering requirements for this robot are discussed, as are the evolution of those requirements into characteristics and then into detail designs. Mechanism designs are presented for the unique modular rod-ends used in SUPERball, including the cable routing system and actuator. Custom force gauges are developed, evaluated, and used to show proof-of-concept sensing on SUPERball. Structural testing is performed to evaluate SUPERball’s rod-end housing, and future improvements are discussed based on all these observations. Finally, preliminary locomotion is shown, using the fully constructed SUPERball robot from these designs.
Light refreshments will be provided!
A Statistical Learning Approach for Detection of a Chiller Energy Efficiency Fault, Lily Hu, Wed. August 15 pm, 2:00pm, 6101 Etcheverry Hall
Abstract: Commercial buildings consume 19% of US primary energy. Of this, an estimated 15% to 30% of energy used in commercial buildings is wasted by poorly maintained, degraded, and improperly controlled equipment. To detect degradation in the energy efficiency of a chiller in a chiller plant, a multivariate Gaussian mixture model is applied. This classification technique was selected due to hypothesized correlation between variables because of their physical interpretation and centering on a mean due to equipment and operation specifications and system control targets. The hidden variable is the faultiness of the chiller and can take on one of three possibly states. The five observed variables correspond to sensor measurements that are commonly available on commercially available chillers and that are commonly monitored in commercial chiller plants. The fault detection algorithm is trained on simulated data for the Molecular Foundry at the Lawrence Berkeley National Laboratory. Then, the model is tested on measured sensor data. The results show that detection of a severe fault and of no fault is the easiest, while detection of moderate faults is sometimes mistaken for severe faults. Further analysis on sensitivity to training data and the fit of the probability distribution is also conducted. The computation needs are moderate enough for deployment and continuous energy monitoring.
Learning about Learning and Engineering: Engineers, Students, and Educators Co-Design Challenges for a Science Center, Jennifer Wang, Wed. June 11, 1 pm, Berkelely Institute of Design
Abstract: We present two case studies of cross-community collaborations of museum educators, engineers from industry, and undergraduate engineering students tasked with co-designing engineering challenges for a science center’s drop-in engineering tinkering program. Each collaboration worked over a semester to research, brainstorm, design, develop, implement, and refine design challenges that represent authentic design practices of the collaboration’s industry engineers. The first collaboration involved engineering students from an education outreach club along with engineers from a software company, and the second collaboration involved engineering students from a product development course along with engineers from a sound reinforcement company.
Collaborative Design Infomatics: Leveraging Data to Make Design Teams Better, Mark Fuge, Mon. June 9, 3 pm, Berkelely Institute of Design
Abstract: The nature of product design is evolving, both inside corporations and in self-organized online communities (e.g., OpenIDEO, VehicleForge). This is thanks to unprecedented amounts of digital design information made possible by globally distributed groups of thousands of people who collaborate together on design projects over the Internet. However, this plethora of information comes with a price: individuals cannot process all of it in a reasonable time frame, limiting their potential.
In this talk I will describe how to apply machine learning techniques to help designers navigate and use this vast quantity of information. Specifically, I will present a story about a particular design community, OpenIDEO, and some of the challenges they face: How do you maintain a sustainable and creative design community without centralized command? How do designers locate the most relevant or creative inspirations out of thousands of ideas? How do novice designers use the community to learn what design methods are appropriate for a given problem? Framing these real-world problems through the lens of Network Analysis, the Maximum Coverage Problem, Link Prediction, and Recommender Systems, I will summarize the empirical performance modern algorithms achieve in practice, describe the major stumbling blocks that need to be overcome, and present algorithms that ameliorate some of those issues. In addition, I will discuss the implications my results have on what role data-driven techniques should play in the creative design of new products.
Engineering Learning: Cross-Community Design, Development, and Implementation of Engineering Design Challenges at a Science Center, Jennifer Wang, Wed., June 4, Noon, Lawrence Hall of Science
Abstract: What have I been doing for the past 3+ years besides just hanging out at the Hall? And why have I been randomly videotaping visitors and asking them questions? Who are all those strangers that I’ve brought to the Hall? In this talk, I will summarize my dissertation research at the Hall, specifically at the Ingenuity Lab. My research develops a deeper understanding of tinkering spaces at public science centers as an accessible pathway towards engineering. I engage in cross-community collaborations with college engineering students, industry engineers, and informal science educators to design challenges for the Ingenuity Lab, and I analyze the effect of these challenges on the learner experience. My dissertation explores these collaborators’ processes of developing engineering design challenges as well as the engineering design processes in which visitors engage at the Ingenuity Lab.
Reducing Work-Related Joint Pain, Christina Ashley Yee, Fri. May 9, Noon, BEST Lab, 230 Hesse Hall
Practice Quals for Christina Yee.
Abstract: Workers across a wide variety of occupations experience work-related joint pain. While some occupations lack adequate solutions to work-related injuries, other fields’ solutions are not feasible for workers due to cost and/or functionality. Therefore, the overall goal of my research is to reduce occupational injuries by designing assistive devices which provide comfortable joint support using passive technology to reduce costs. In particular, I will focus on two projects: 1) reducing dentists’ neck pain when performing dental procedures and 2) reducing construction workers’ lower back pain when moving materials.
Valuing Design, Alice M. Agogino, Fri. May 2, 2:30-3:30 pm B100 Blum Hall
BEST Lab Director Alice Agogino will be giving a talk for the Spring Seminar on Design and Innovation.
Abstract: Quality design is credited with providing a powerful competitive advantage to businesses and driving successful innovation. Metrics for measuring the value of design have included performance ratings (e.g., J.D. Power, Consumer Reports), financial metrics (e.g., stock performance, revenues, sales), expert evaluations (e.g., IDEA/Business Week, Consumer Electronics Society awards), creativity metrics (e.g., novelty, variety), process metrics (e.g., multidisciplinary teamwork, human-centered design), and societal impact (e.g., environmental, development in emerging regions). A survey of research in selected design metrics will be presented along with a discussion of what metrics might be used to differentiate UC Berkeley’s Design Innovation program. What skills do we hope our design graduates will have when they leave UC Berkeley? What role will our students play in creating a more meaningful and sustainable world? In this talk, design will be viewed broadly to include the design of physical products, software, services, experiences, new business models and policy.
How We Design: From Animal Innovation to Engineering Innovations, Andy Dong, Thu. April 24, 4-5:00 pm 290 Hearst Mining
Former BEST Labber Andy Dong will be giving a talk for the Spring Seminar on Design and Innovation. He is currently Warren Centre Chair for Engineering Innovation at the University of Sydney, Australia.
Abstract: Humans have a nearly unbounded capability to design our environment to suit our current and future needs, mostly for the better but sometimes with injurious unintended consequences. The products of our capability to design are all around us, and we are never apparently satisfied with what we have imagined and made, always pursuing something better. Where does this capability to design come from and why have we not seen evidence of this capability in nonhuman species and artificial systems? In this talk, I will explore the cognitive skills and cognitive strategies we have adapted, possibly optimally, to engineer and expand environments within which we could survive. I will look into the available evidence from comparative psychology on animal innovation and from cognitive design research on the cognitive strategies of designers. I will then explain how these cognitive skills and strategies have provided a basis for my research. Specifically, I will discuss their influence on 1) engineering design tools and 2) a normative, ethical framework for design based on the capability approach pioneered by Nobel Laureate Amartya Sen and legal ethicist Martha Nussbaum.
Thursday, March 20th, 4:00 pm, BEST Lab (230 Hesse Hall): Jennifer Wang “Design Challenges at a Science Center: Are Children Engineering?”
Abstract: Engineering design challenges and tinkering activities are increasingly popular in informal learning settings. Thus, these environments can benefit from foundational research on learners’ engineering processes as they engage in these settings. This paper conducts an exploitive analysis of an existing design challenges program at a science center through the investigation of the design processes of 22 visitor groups across five challenges. The design processes are compared across the challenges to identify characteristics of these settings that engage learners in engineering. The premise, example designs, and materials played key roles in visitors’ design processes. Findings show that (1) each challenge provided unique contexts in which to engage in iterative engineering design, (2) visitors utilized existing designs and designs in progress from other visitors for inspiration, and (3) visitors were particularly influenced by the materials and used them as a means to gather information, explore possibilities, and identify goals. Many visitors also exhibited design process progressions similar to expert engineers, suggesting that the context and materials provide opportunities for early engineering experiences.
Friday, March 14th, 12-1:00 pm, BEST Lab (230 Hesse Hall): Euiyoung Kim “Developing (Human-Centered) Design Roadmap as a means of setting a firm’s long term strategy: Connecting dots of products, technologies, and design roadmaps”
Abstract: Product roadmapping is a useful method for enterprises to keep their product strategies up-to-date and to predict upcoming market trends faster and easier ways by illustrating the visualization of product portfolios over time-frame. Firms keep it updated in a regular time base and use it for the guideline to decide which products and technologies should be selected and funded for the next product/service development. However, due to the misalignment between technologies on hand and the product roadmap firms expect to have, they easily get into trouble to coupling products and technologies in a way to leverage core competences together, which threaten overall company’s competitiveness and fundamentally lead to the failure of the market forecast. This research proposes an integrated human-centered design roadmap that helps companies in new product development industry to predict future products in accurate ways and to have a sustainable new product/service development strategy.
Friday, February 14th, 12-1:00 pm, BEST Lab (230 Hesse Hall): Mark Fuge “Collaborative Design Informatics”
Abstract: The nature of product design is evolving, both inside corporations and in self-organized online communities (e.g., OpenIDEO, VehicleForge). This is thanks to unprecedented amounts of digital design information made possible by globally distributed groups of thousands of people who collaborate together on design projects over the Internet. However, this plethora of information comes with a price: individuals cannot process all of it in a reasonable time frame, limiting their potential.
In this talk I will describe how to apply machine learning techniques to help designers navigate and use this vast quantity of information. Specifically, I will present a story about a particular design community, OpenIDEO, and some of the challenges they face: How do you maintain a sustainable and creative design community without centralized command? How do designers locate the most relevant or creative inspirations out of thousands of ideas? How do novice designers use the community to learn what design methods are appropriate for a given problem? Framing these real-world problems through the lens of Network Analysis, the Maximum Coverage Problem, Link Prediction, and Recommender Systems, I will summarize the empirical performance modern algorithms achieve in practice, describe the major stumbling blocks that need to be overcome, and present algorithms that ameliorate some of those issues. In addition, I will discuss the implications my results have on what role data-driven techniques should play in the creative design of new products.