Abstract : Corresponding to the organizational efforts introduced in the mini-symposium “Frontiers of Collaboration with Industry: Towards International Mathematical Commons,” this mini-symposium introduces diverse efforts at practical research activities that engage mathematicians, together with companies, to solve specific problems. With a spirit of collaboration, we aim at sharing with participants some issues and challenges at the forefront of mathematical science research, a collaborative system with companies, and various mathematical ideas for application to problem-solving. Through collaboration between industry and mathematical science, all are intended for the successful resolution of the many issues which must be addressed by our society.
[01441] Mobility Optimization Engine and its Real-world Applications
Format : Talk at Waseda University
Author(s) :
Katsuki Fujisawa (Kyushu University)
Abstract : Various efforts have been made to realize a so-called super-smart society recently. Our project team builds services to create new industries and other services through corporate collaboration. We have utilized large-scale computing infrastructures and developed the Cyber-Physical System Mobility Optimization Engine (CPS-MOE) that provides various functions, including creating new industries. It can reduce cost and industrial waste and constructing services to calculate the optimum control schedule of transportation agencies. The latest research results and industry-academia collaborative projects using CPS-MOE will be presented in this talk.
[02078] Mathematical modeling with industry in the water sector: what makes good practice
Format : Talk at Waseda University
Author(s) :
Anthony John Jakeman (Australian National University)
Abstract : We emphasize the role of good practices in conducting an integrated assessment exercise in water availability settings, underlining attention throughout the framing, assessment and engagement steps. We stress the notion of reflexivity on pathway decisions at each decision fork in the exercise and a holistic attention to uncertainty sources throughout the process, not just in the formulated models. These aspects will be illustrated with a case study in a catchment of the Murray-Darling Basin.
[01474] An international research program on industrial problems for math students.
Format : Talk at Waseda University
Author(s) :
Hiroshi Suito (Tohoku University)
Abstract : In this talk, a student research program with industrial projects in mathematics, called “g-RIPS-Sendai”, which has been held in Japan since 2018 is introduced. This program offers graduate students in mathematics stimulating opportunities to work on realistic research projects provided by industries. For industrial partners, this program provides chances to try new mathematical approaches with fresh ideas from young students. Our experiences including several difficulties will be shared and discussed.
[02098] Collaboration with early graduate researchers, and improvements on simulated annealing
Format : Talk at Waseda University
Author(s) :
Joseph David (University of Washington)
Kemal Aziez Rachmansyah (Tohoku University)
Rikuto Shigemi (University of Tsukuba)
Zachary Brennan (Iowa State University)
Abstract : Students with a pure mathematics background can struggle to find early opportunities in industry without the correct guidance. As opposed to traditional graduate internships which may expect certain familiarity with industry-related problems, a hybrid academic-industry approach with both academic and industry mentors can provide a bridge for early graduate students wishing to transition to industry. This talk discusses the pros and cons of one such program from the perspective of a mid-career graduate student.
[02914] Development of a CNN-Based Model for Car Classification and Damage Detection from car accidents
Format : Talk at Waseda University
Author(s) :
Busayamas pimpunchat (King Mongkut’s Institute of Technology Ladkrabang)
Abstract : This study aimed to develop a convolutional neural network (CNN)-based model for car classification and damage detection from car accidents. The research found that the model was highly accurate in classifying the position of a single car and detecting damage and the model's performance in detecting multiple cars, a modified version of the Hough Transform algorithm was used to improve. The study's findings can help insurance companies analyze car damage and introduce relevant sections in vehicle regulations.
[03805] Failure and Success for Bauxite Moisture Measurement with Microwaves
Format : Talk at Waseda University
Author(s) :
Mark Joseph McGuinness (Victoria University of Wellington)
Lata I Paea (University of the South Pacific)
Sione Paea (University of the South Pacific)
Abstract : The weight of bauxite ore is strongly dependent on moisture content. A microwave analyser measures phase and attenuation in real time, to infer moisture content, which directly impacts the price paid per tonne. Simple linear models fail to explain the highly nonlinear dependence on bauxite depth that is seen in attenuation data. Motivated by this failure, we develop a successful four-layer model that allows multiple reflections at interfaces between ore and air.
[02161] The UniSA Mathematics Clinic: Renewables, defence, mining, drones, and justice
Format : Talk at Waseda University
Author(s) :
Lesley Ann Ward (University of South Australia)
Abstract : The University of South Australia Mathematics Clinic runs yearlong authentic industry-sponsored research projects for small teams of final-year undergraduate mathematics students. Clinic projects give students professional skills, experience, accomplishments, and employability, and generate research progress and publications. But things can go wrong – with sponsors, students, staff, projects, workload, and logistics. How can we minimise things going wrong, and fix them when they do? I’ll offer some case studies, suggestions, and principles, drawn from Clinic experience.
[01470] Interdisciplinary research using Topological Flow Data Analysis through Math Clinic
Format : Talk at Waseda University
Author(s) :
Takashi Sakajo (Kyoto University)
Abstract : Topological Flow Data Analysis "TFDA" is a new way of data analysis based on the classification theory of 2D vector fields. It extracts and quantifies topological features of complex flows from various datasets efficiently. We are organizing a research platform, called Math Clinic, at Kyoto University, where interdisciplinary research with other disciplines including industry has been conducted based on TFDA. In this talk, we will report on the activity and results of the Math Clinic.
