Associate Professor, director of the Smart Mobility Analytics Research and Training Lab Industrial & Manufacturing System Engineering University of Michigan - Dearborn 4901 Evergreen Rd. Dearborn, MI 48128. Phone: (313) 583-6745 Fax: (313) 593-3692 E-mail: jianhu@umich.edu
Dr. Hu is an Associate Professor in the Department of Industrial & Manufacturing Systems Engineering at the University of Michigan - Dearborn. He received his PhD in the area of operations research from Northwestern University 2011, M.S. in the area of control theory and applications from the University of Arkansas at Little Rock and the Chinese Academy of Sciences, and B.S. from the Xi'an Jiaotong University. Dr. Hu's research interests focus on methodologies of data-driven risk management and decision making uncertainty and their applications including
He has led and participated in numerous federally and industrially funded research projects on robust assessment of consumer preference, advanced robust optimization models and algorithms, driving behavior data analytics, emergency management of traffic incidents, risk-adjusted homeland security budget allocation, resilient nurse scheduling and rescheduling, and plant workforce planning.
Robust averse machine learning driven project management
This project is to pursue the best portfolio investment among vehicle development projects for enhancing vehicle performance, economy, and after-sales service.
Plant workforce planning optimization (collaboration with Ford)
This project is to find an optimal quantity of legacy and TPT employees for minimizing staff cost and controlling the risk that staff headcounts cannot meet the rise and fall in production demand for a long-term planning horizon.
Deployment of public charging infrastructure (collaboration with the Argonne National Laboratory)
This project is to assess EV drivers’ charging preferences, predict charging demands, and on this basis design an optimal long-term plan for deploying public charging infrastructure and upgrading current power distribution networks.
Counterfactual explanation for machine learning
This project is to interpret and explain the predictions of individual instance given by machine learning models.
Robust decision making of coordinating pricing and inventory control (funded by the NSF)
TThis project develops a data-driven robust model to seek the best-selling price and order quantity for maximizing a retailer’s profit when facing the challenge that there is a lack of market data to accurately predict the price-demand function.
Risk adjusted homeland security budget allocation (funded by the NSF)
This project is to model a budget allocation problem for the Urban Areas Security Initiative (UASI) against property losses, fatalities, and damages of important infrastructure in terrorist attacks.
Risk averse multi-attribute decision making based on stochastic dominance (funded by the NSF)
This project is to develop a novel risk-averse decision making framework which maximizes the profit while requiring the decision policy preferable to the benchmark to the extent that the random consequence yielded by the policy stochastically dominated one at the benchmark.
Traffic incident response and management (funded by the FHWA)
This project is to develop a traffic incident response and management model for seeking the best strategy of dispatching service vehicles in Central Arkansas to fast handle existing traffic incidents and well prepare for potential ones.
Human driving behaviors and traffic waves (collaboration with Ford)
This project is to demonstrate a bifurcation perspective of the discrete-time dynamical system in a car-following context.
Resilient nurse rescheduling model under demand uncertainty and labor absenteeism (collaboration with Henry Ford Macomb Hospital)
This project is to find an optimal policy for both maintaining the quality of patient care and respecting the welfare of nurses. A crucial problem worthy of consideration is uncertainty on a daily basis due to emergency responses for patients and absenteeism among registered nurses. The uncertainty leads to an unforeseen shortage of nursing workload, and as a result, decreases standards of care.
Effect of production and consumption taxes on emission control under market uncertainty
This research proposes a game theory based model for evaluating and comparing the effect of production and consumption taxes on promoting the usage of new technology and controlling emission under market uncertainty and instability
Car following model using V2V communication-based technology
This research models homogeneous car following behavior with information sharing between vehicles using advanced V2V communication-based technology. This information sharing assists drivers’ interactions and promotes safety pilots.
Papers Published or Accepted for Publication in Refereed Journals
Conference Proceedings
Courses taught at the University of Michigan Dearborn
Current graduate students
PhD students
Former graduate students
Societies
Stochastic Programming pages:
Stochastic Programming pages:
(313) 583-6745
(313) 593-3692
jianhu@umich.edu