Mohammad Hajiesmaili
Associate Professor of Computer Science University of Massachusetts Amherst
- Amherst MA
Computer scientist Mohammad Hajiesmaili researches algorithms, sustainable computing and data center energy optimization.
University of Massachusetts Amherst
View more experts managed by University of Massachusetts Amherst
Expertise
Biography
Dr. Hajiesmaili's honors include an NSF CAREER Award and recognition in Popular Science’s Brilliant 10 in 2022.
Video
Education
University of Tehran
Ph.D.
Electrical and Computer Engineering
University of Tehran
M.Sc.
Electrical and Computer Engineering
Sharif University of Technology
B.Sc.
Computer Engineering
Select Recent Media Coverage
Decarbonizing the Internet
Popular Science Online
Popular Science’s Brilliant 10 feature covers research approaches to reducing data centers’ carbon footprint, including carbon-aware computing and energy storage. Hajiesmaili’s work on reducing computing’s carbon footprint is part of the feature.
Select Publications
Toward Environmentally Equitable AI
Communications of the ACMMohammad Hajiesmaili, Shaolei Ren, Ramesh Sitaraman, Adam Wierman
2025-07-01
Dr. Hajiesmaili and colleagues advocate for environmental equity as a priority for the management of future globally deployed AI systems. Concretely, they explore the potential of harnessing AI workloads' scheduling flexibility and using equity-aware GLB as a lever to fairly redistribute the environmental cost across regions, ensuring that no single region disproportionately bears the environmental burden. Then, they present key algorithmic challenges to enable AI's environmental equity without significantly degrading the other performance metrics, such as energy cost and inference accuracy. Finally, they discuss future directions to unleash the full potential of system management for environmentally equitable AI, including coordinated scheduling of AI training and inference, joint optimization of IT and non-IT resources, holistic control of system knobs, and building theoretical foundations. Their proposal of environmental equity advances the boundaries of existing research on sustainable AI and mitigates the otherwise uneven distribution of AI's environmental costs across different regions.
Online EV Charging Scheduling With On-Arrival Commitment
IEEE Transactions on Intelligent Transportation SystemsBahram Alinia, Mohammad Hajiesmaili, Noël Crespi
2019-01-11
Dr. Hajiesmaili and colleagues develop online schedules for electric vehicle charging when station capacity is limited and vehicle arrivals are uncertain. Their algorithms commit to a guaranteed amount of charging on arrival and are designed to encourage truthful reporting; simulations using real traces compare them with an optimal offline approach without arrival-time commitments.
Incentivizing Device-to-Device Load Balancing for Cellular Networks: An Online Auction Design
IEEE Journal on Selected Areas in CommunicationsMohammad Hajiesmaili, Lei Deng, Minghua Chen, Zongpeng Li
2017-02-01
Dr. Hajiesmaili and colleagues design an online auction to pay devices to relay traffic from busy cellular areas to nearby underused ones, helping balance network load. The auction makes truthful bidding a dominant strategy for devices; tests using real-world traces show costs 45% lower than with an alternative heuristic.
Peak-Aware Online Economic Dispatching for Microgrids
IEEE Transactions on Smart GridYing Zhang, Mohammad Hajiesmaili, Sinan Cai, Minghua Chen, Qi Zhu
2016-04-06
Dr. Hajiesmaili and colleagues study online economic dispatch for microgrids, which schedule local power sources to meet demand amid uncertainty in demand and renewable generation. Their peak-aware algorithms account for peak-based electricity pricing and perform close to an offline optimum in evaluations using real-world traces.
