About

Iman Soltani is a professor affiliated with the Laboratory for AI, Robotics, and Automation (LARA) at the University of California, Davis. His research focuses on developing autonomous systems that learn, adapt, and operate in complex environments. This includes creating intelligent systems capable of independent operation in challenging settings, with applications spanning industrial diagnostics to enhance manufacturing efficiency and safety, as well as health diagnostics aimed at improving medical diagnostics and healthcare delivery through AI and robotics. Professor Soltani's work integrates advanced machine learning techniques with robotics to address real-world problems, contributing to the fields of autonomous navigation, robotic manipulation, and assistive technologies. His research outputs include contributions to conferences and journals in machine learning, robotics, automation, and medical imaging, reflecting a broad interdisciplinary approach to advancing autonomous systems and their practical applications.

Research topics

• Computer Science
• Machine Learning
• Artificial Intelligence
• Data Mining
• Mathematics

Selected publications

• Flexible Operation and Control of Bidirectional Chargers Using ISMC for Electric Vehicle Applications

  Journal of Sustainability Research · 2026-03-09

  articleOpen access1st authorCorresponding

  The increasing integration of electric vehicles and renewable energy systems requires high-performance, reliable, and intelligent battery charging infrastructure. This paper presents a novel design and analysis of a bidirectional fast battery charging system connected to the power grid, employing a robust improved sliding mode control (ISMC) to overcome the limitations of conventional control methods. Conventional controllers such as PID, FL-PID, and ADRC suffer from parameter sensitivity, computational burden, and noise susceptibility under uncertain and dynamic conditions. In contrast, the proposed ISMC approach achieves improved robustness, fast dynamic response, and high tracking accuracy. The proposed architecture consists of two stages: a three-phase bidirectional voltage source converter on the AC side and a buck-boost DC/DC converter on the DC side. The voltage source converter employs the proposed ISMC to regulate the dq-axis grid currents and manage the active/reactive power flow, while the DC/DC converter is also governed by an ISMC approach to stabilize the DC-link and battery voltage across varying charge/discharge states. A significant contribution of this work is the proposed flexible operating algorithm that enables seamless transitions between grid-connected and islanded modes, ensuring uninterrupted supply to local AC and DC loads during grid outages through grid-forming control. Additionally, a battery management algorithm is integrated, which dynamically adjusts charging and discharging actions based on the battery’s state of charge, thereby prolonging battery lifespan and ensuring safe operation. Verifications via simulations in MATLAB/Simulink confirm the superiority of the proposed control strategy by demonstrating less than 3% total harmonic distortion in grid current, accurate current tracking, and a stable DC voltage profile with negligible overshoot and fast settling time. In comparison, the PI controller exhibits 0.4% overshoot and 50 ms settling time, while the ADRC shows about 0.3% overshoot and 40 ms settling time. The ISMC-based framework can be applied to next-generation smart battery charging stations, providing enhanced grid support, improved power quality, and advanced battery health management.

  PublisherOA PDFDOI

• VITA: Vision-to-Action Flow Matching Policy

  ArXiv.org · 2025-07-17

  preprintOpen accessSenior author

  Conventional flow matching and diffusion-based policies sample via iterative denoising from standard noise distributions (e.g., Gaussian), and require conditioning modules to repeatedly incorporate visual information during the generative process, incurring substantial time and memory overhead. To reduce the complexity, we develop VITA, VIsion-To-Action policy, a noise-free and conditioning-free flow matching policy learning framework that directly flows from visual representations to latent actions. Since the source of the flow is visually grounded, VITA eliminates the need for visual conditioning during generation. As expected, bridging vision and action is challenging, because actions are lower-dimensional, less structured, and sparser than visual representations; moreover, flow matching requires the source and target to have the same dimensionality. To overcome this, we introduce an action autoencoder that maps raw actions into a structured latent space aligned with visual latents, trained jointly with flow matching. To further prevent latent action space collapse during end-to-end training, we propose flow latent decoding, which anchors the latent generation process by backpropagating the action reconstruction loss through the flow matching ODE (ordinary differential equation) solving steps. We evaluate VITA on 9 simulation and 5 real-world tasks from ALOHA and Robomimic. VITA achieves 1.5x-2x faster inference compared to conventional methods with conditioning modules, while outperforming or matching state-of-the-art policies. Project page: https://ucd-dare.github.io/VITA/.

  PublisherOA PDFDOI

• Design and Implementation of a Dual Uncrewed Surface Vessel Platform for Bathymetry Research under High-flow Conditions

  arXiv (Cornell University) · 2025-02-18

  preprintOpen accessSenior author

  Bathymetry, the study of underwater topography, relies on sonar mapping of submerged structures. These measurements, critical for infrastructure health monitoring, often require expensive instrumentation. The high financial risk associated with sensor damage or vessel loss creates a reluctance to deploy uncrewed surface vessels (USVs) for bathymetry. However, the crewed-boat bathymetry operations, are costly, pose hazards to personnel, and frequently fail to achieve the stable conditions necessary for bathymetry data collection, especially under high currents. Further research is essential to advance autonomous control, navigation, and data processing technologies, with a particular focus on bathymetry. There is a notable lack of accessible hardware platforms that allow for integrated research in both bathymetry-focused autonomous control and navigation, as well as data evaluation and processing. This paper addresses this gap through the design and implementation of two complementary USV systems tailored for uncrewed bathymetry research. This includes a low-cost USV for Navigation And Control research (NAC-USV) and a second, high-end USV equipped with a high-resolution multi-beam sonar and the associated hardware for Bathymetry data quality Evaluation and Post-processing research (BEP-USV). The NAC-USV facilitates the investigation of autonomous, fail-safe navigation and control, emphasizing the stability requirements for high-quality bathymetry data collection while minimizing the risk to equipment. The BEP-USV, which mirrors the NAC-USV hardware, is then used for additional control validation and in-depth exploration of bathymetry data evaluation and post-processing methodologies. We detail the design and implementation of both systems, and open source the design. Furthermore, we demonstrate the system's effectiveness in a range of operational scenarios.

  PublisherOA PDFDOI

• A 2.69-ppm/°C curvature-compensated BJT-based bandgap voltage reference

  Integration · 2025-01-22 · 2 citations

  articleOpen accessCorresponding

  This research presents a BJT-based bandgap reference circuit, aiming to minimize the temperature coefficient and active area for low-power and compact applications. A curvature compensation technique is introduced to enhance the temperature coefficient and extend the operational temperature range. The proposed BGR, simulated using a 0.18-μm CMOS process, demonstrates a simulated reference voltage of 0.269 V and TC of 2.69 ppm/°C for the reference output across a wide temperature range of −50 °C–150 °C. Furthermore, the proposed circuit occupies a compact silicon area of 0.0054 mm 2 , shows a line regulation 0.46 %/V, and consumes a power of 22.07 μW at 25 °C. The proposed bandgap reference circuit well-suited for providing reference voltages in various integrated circuits, particularly in high-precision low-power applications. In this research a current-mode BGR based on three BJT transistors, a resistor, and MOS transistors is proposed. In compared with the prior BGR. • This research introduces a current-mode dual-reference BGR featuring process-insensitive piecewise curvature compensation, which minimizes nonlinearity across a broad temperature range. • To achieve a reduction in chip area to less than 0.007 mm 2 (smaller area than the previous BGR), this paper proposes the elimination of output passive resistance, and the circuit employs the gate-source voltage of MOS transistor to generate reference voltage. • Additionally, it introduces an approach by utilizing just an operational amplifier to generate PTAT and CTAT voltages that differs from those employed in most prior structures. • In this design, the CTAT voltage is generated using a PMOS transistor in conjunction with a PNP BJT transistor, which is biased by a PTAT voltage. This methodology (biasing by a PTAT voltage) enhances the temperature coefficient of the CTAT current, thereby minimizing output variations. • This design incorporates a curvature compensation circuit to expand the temperature ranges, which first, significantly decreases the number of transistors in the compensation circuit from previous configurations, and second, it can work in subthreshold region to reduce the power consumption. • The power consumption, temperature range, area, and minimum supply voltage are improved in comparison with the previous works. • The proposed circuit potentially offers greater flexibility in providing multiple outputs while maintaining low-power and low-voltage characteristics.

  PublisherDOI

• A Regulatory-Driven Planning Framework for Fast Charging Station Deployment in Power Distribution Systems

  SSRN Electronic Journal · 2025-01-01

  preprintOpen access1st authorCorresponding
  PublisherDOI

• User-Centered Insights into Assistive Navigation Technologies for Individuals with Visual Impairment

  ArXiv.org · 2025-04-08

  preprintOpen access1st authorCorresponding

  Navigational challenges significantly impact the independence and mobility of Individuals with Visual Impairment (IVI). While numerous assistive technologies exist, their adoption remains limited due to usability challenges, financial constraints, and a lack of alignment with user needs. This study employs a mixed-methods approach, combining structured surveys and virtual workshops with 19 IVI to investigate their experiences, needs, and preferences regarding assistive technologies for navigation and daily living. The survey results provide insights into participants technological competence, preferences for assistive devices, and willingness to adopt new solutions. In parallel, workshop discussions offer qualitative perspectives on key navigation challenges, including difficulties in detecting overhead obstacles, navigating environments with complex layout, and the limitations of existing technologies. Findings highlight the need for assistive devices that integrate both navigational guidance and high-level spatial awareness, allowing users to build mental maps of their surroundings. Additionally, multimodal feedback, combining audio, haptic, and tactile cues, emerges as a crucial feature to accommodate diverse user preferences and environmental conditions. The study also underscores financial and training barriers that limit access to advanced assistive technologies. Based on these insights, we recommend the development of customizable, user-friendly, and most importantly affordable navigation aids that align with the daily needs of IVI. The findings from this study provide guidance for technology developers, researchers, and policymakers working toward more inclusive and effective assistive solutions.

  PublisherOA PDFDOI

• Look, Focus, Act: Efficient and Robust Robot Learning via Human Gaze and Foveated Vision Transformers

  ArXiv.org · 2025-07-21

  preprintOpen accessSenior author

  Human vision is a highly active process driven by gaze, which directs attention to task-relevant regions through foveation, dramatically reducing visual processing. In contrast, robot learning systems typically rely on passive, uniform processing of raw camera images. In this work, we explore how incorporating human-like active gaze into robotic policies can enhance efficiency and robustness. We develop GIAVA (Gaze Integrated Active-Vision ALOHA), a robot vision system that emulates human head and neck movement, and gaze adjustment for foveated processing. Extending the AV-ALOHA robot platform, we introduce a framework for simultaneously collecting eye-tracking, perspective control, and robot manipulation demonstration data from a human operator. We also open-source a simulation benchmark and dataset for training robot policies that incorporate human gaze. Inspired by recent work in foveated image segmentation and given the widespread use of Vision Transformers (ViTs) in robot learning, we integrate gaze information into ViTs using a foveated patch tokenization scheme. Compared to uniform patch tokenization, this significantly reduces the number of tokens, and thus computation. Our results show that our method for foveated robot vision drastically reduces computational overhead, and enhances robustness to background distractors. Notably, on certain high-precision tasks, foveated vision also improves performance, as reflected in higher success rates. Together, these findings suggest that human-inspired foveated visual processing offers untapped potential and should be further considered as a useful inductive bias in robotic vision systems. https://ian-chuang.github.io/gaze-av-aloha/

  PublisherOA PDFDOI

• Design and simulation of vehicle motion tracking system using a Youla controller output observation system

  Measurement and Control · 2025-06-17

  articleOpen accessSenior author

  This paper presents a novel linear robust Youla controller output observation system for tracking vehicle motion trajectories using a simple nonlinear kinematic vehicle model, supplemented with positional data from a radar sensor. The proposed system operates across the full vehicle trajectory range with only three linear observers, improving upon previous methods that required four nonlinear observers. To ensure smooth transitions between Youla controllers and observers, a switching technique is introduced, preventing bumps during controller changes. The proposed observer system is evaluated through simulations, demonstrating accurate and robust estimation of longitudinal and lateral positions, vehicle orientation, and velocity from sensor measurements during various standard driving maneuvers. Results are provided for different driving scenarios, including lane changes and intersection crossings, where significant changes in vehicle orientation occur. The novelty of this work lies in the first application of a Youla controller output observer for vehicle tracking estimation.

  PublisherDOI

• Active Vision Might Be All You Need: Exploring Active Vision in Bimanual Robotic Manipulation

  2025-05-19 · 4 citations

  articleSenior author

  Imitation learning has demonstrated significant potential in performing high-precision manipulation tasks using visual feedback. However, it is common practice in imitation learning for cameras to be fixed in place, resulting in issues like occlusion and limited field of view. Furthermore, cameras are often placed in broad, general locations, without an effective viewpoint specific to the robot's task. In this work, we investigate the utility of active vision (AV) for imitation learning and manipulation, in which, in addition to the manipulation policy, the robot learns an AV policy from human demonstrations to dynamically change the robot's camera viewpoint to obtain better information about its environment and the given task. We introduce AV-ALOHA, a new bimanual teleoperation robot system with AV, an extension of the ALOHA 2 robot system, incorporating an additional 7-DoF robot arm that only carries a stereo camera and is solely tasked with finding the best viewpoint. This camera streams stereo video to an operator wearing a virtual reality (VR) headset, allowing the operator to control the camera pose using head and body movements. The system provides an immersive teleoperation experience, with bimanual first-person control, enabling the operator to dynamically explore and search the scene and simultaneously interact with the environment. We conduct imitation learning experiments of our system both in real-world and in simulation, across a variety of tasks that emphasize viewpoint planning. Our results demonstrate the effectiveness of human-guided AV for imitation learning, showing significant improvements over fixed cameras in tasks with limited visibility. Project website: https://soltanilara.github.io/av-alohal

  PublisherDOI

• Design and Analysis of a Novel Two-Switch Transformerless High Step-Up DC–DC Converter for Grid-Connected Renewable Energy Sources

  IEEE Access · 2025-01-01 · 1 citations

  articleOpen access1st authorCorresponding

  Renewable Energy Sources (RESs), particularly Photovoltaic (PV) systems, inherently produce variable DC voltages that often cannot meet load or grid requirements directly. Consequently, a reliable and efficient DC-DC converter is required to interface RESs with downstream converters and loads. This paper proposes a non-isolated, high-gain, transformerless step-up DC-DC converter that provides continuous input current and a non-inverting output voltage. The proposed topology is designed to minimize switching losses and to maintain a low component count, thereby improving conversion efficiency while containing cost and implementation complexity. A comprehensive analytical model that includes parasitic elements is developed to derive the converter voltage gain. Comparative analyses of device voltage and current stresses against recently reported topologies are presented. The results demonstrate that the proposed converter achieves high voltage gain with reduced voltage and current stresses on switching devices, while preserving acceptable component current levels. Conduction intervals of switches and diodes, as well as switching loss contributions, are analyzed and quantified. Experimental validation is provided by a 100 W prototype, and measured results corroborate the theoretical predictions and simulation outcomes. The proposed converter thus represents an effective and practical solution for high-gain DC–DC conversion in renewable energy applications, offering an advantageous trade-off between efficiency, component simplicity, and cost.

  PublisherDOI

Frequent coauthors

• Ming Liang

  Changchun University of Science and Technology

  13 shared

• Kamal Youcef‐Toumi

  Massachusetts Institute of Technology

  13 shared

• Mohammad Sarvi

  Iran University of Science and Technology

  9 shared

• Eric Darve

  6 shared

• Ziyi Yang

  Yunnan Agricultural University

  5 shared

• Amin Ghafourian

  University of California, Davis

  5 shared

• Andreas Schuh

  4 shared

• Ivo W. Rangelow

  Technische Universität Ilmenau

  4 shared

Labs

• Laboratory for AI, Robotics, and Automation (LARA)PI

  Developing autonomous systems that learn, adapt, and operate in complex environments at UC Davis.

Education

• Doctor of Philosophy , Mechanical Engineering

  Massachusetts Institute of Technology

  2015

• Master of Applied Science , Mechanical Engineering

  University of Ottawa

  2007

• Bachelor of Science, Mechanical Engineering

  Amirkabir University of Technology

  2003