Nexus Summer 2024 seed grant winners

Mines and NREL formally launched the Nexus initiative in 2019 to strengthen and expand collaboration between the two institutions.

The goal is to develop joint research initiatives, partnerships, student exchanges, and funding opportunities in a way that creates more impact by working together than either institution could do alone.

The Nexus Seed Funding grant program, started in 2020, aims to stimulate new research initiatives and connections across the two institutions, featuring faculty and researchers from Mines and NREL. The Nexus seed funding program has been successful in supporting Mines and NREL researchers in jump-starting ideation on innovative collaborative, interdisciplinary research proposals and prospective funding opportunities (e.g., from DOE, DOD, or other sources).

This year, we received thirteen excellent proposals and have awarded six seed grants to Mines/NREL teams that we hope will lead to joint submissions of compelling collaborative research proposals. The proposals were carefully reviewed by a panel comprised of Mines faculty and NREL researchers to select the 2024 grantees.

 

Mines-NREL Nexus is pleased to announce the following awards for seed funding in the Summer of 2024

Nexus Seed Grants Winners – Summer 2024

 

Human-LLM Loops for Semiconductor Ontology and Knowledge Graph Development

Sagi Zisman (National Renewable Energy Lab), Data Scientist-Computational Scientist and Dr. Gabe Fierro (Colorado School of Mines), Assistant Professor of Computer Science

Semiconductor manufacturing is a complex process and requires massive investments. Recently, the CHIPS and Science Act was passed to ’bolster U.S. leadership in Semiconductors’ with a $52.7 Billion investment including $500 Million for semiconductor supply chains [3]. Specifically, the National Institute of Standards and Technology (NIST) has launched the CHIPS for America program with a few key focal areas [1]. First with a focus on manufacturing digital twins as part of its vision for the CHIPS Manufacturing USA Institute [2]. Secondly, novel metrology techniques for performance and quality assessments of future chips. Thirdly, strengthening the semiconductor supply chain. Just on the supply chain side, from novel architectures, Very Large Scale Integrated (VLSI) design, and analog/digital circuit modeling tools to source materials, fabrication, packaging and distribution, the supply chain of getting a chip or module out to a customer is among the most sophisticated industrial feats humanity has performed. For success in all the above areas, information, knowledge and data management will be critical and indeed is a precondition for effective solutions.

 Our proposal and vision are to focus on research and development of novel knowledge and data management tools that are in direct support of the CHIPS effort but that are ultimately transferable to various domains. Of particular importance to NREL are innovations in Photovoltaic (PV) semiconductors and modules. NREL is increasingly being funded for supply-chain work and given the inherently entity-relational nature of supply chains, knowledge graphs (KGs) are perfect solutions. KGs are graph data structures that encode relationships between interacting entities. Typically, the graph schema is specified as an ontology, or meta graph that describes concepts in the world. What will knowledge graphs enable? These types of questions are being asked constantly in the industry. To answer this question, a multi-hop information nightmare ensues, perhaps requiring merging industry relational databases, internet searches and text-based report synthesis. A query-able KG that links this information will significantly reduce the information friction. Given that supply-chain data contains vast information linking business, geographical, and domain-science, developing a KG is a complex task. To ease the burden, we aim to develop tools for automating the construction of single-source-of-truth KGs. Our vision is to combine state-of-the-art multimodal Large Language Models (LLMs) in a human-LLM interaction loop for the seamless creation and maintenance of large interdisciplinary KGs.

 

Intersection of A.I. and Challenges to Enable Fair, Effective, and Lasting Innovation; An Effort to Develop a Community of Practice and Standards for Prizes & Competitions

Joseph Simon (National Renewable Energy Lab), Researcher V – Systems Engineering and Dr. Sid Saleh (Colorado School of Mines), Associate Professor of Engineering, Design and Society

The role of challenges and competitions to spur innovation, technology commercialization, an impactful energy transition, and a qualified workforce has been well-understood and valued for years. At NREL, the Joint Institute for Strategic Energy Analysis (JISEA) center manages dozens of prize competitions for the U.S. Department of Energy, connecting thousands of innovators with millions of dollars of non-dilutive federal funding that can advance the market in meaningful ways. At Mines, the McNeil Center for Entrepreneurship and Innovation leads efforts to enable students to master the art and science of bringing ideas to life through competing in challenges sponsored by industry partners. Through the America COMPETES act and through extensive private investment, prizes and challenges have been identified as a meaningful way to move the market forward.

NREL & Mines are leaders of how to ensure that the innovation required to achieve a rapid clean energy transition is informed, but not limited by, artificial intelligence systems integrated into competitions and prizes that are fair, effective, and impactful. We will secure new funding to build our community of practice for effective integration of AI into challenges and prizes, informing many federal agencies, state agencies, incubators, accelerators, foundations, and associations who depend on ideas from innovators to advance the state-of-the art. Additionally, the potential benefits of AI applications for prizes should be explored. LLMs can encourage innovators to utilize broader customer research and business model development to directly address pressing societal needs.

 

Augmented Reality Enabled Driving Assistance – the Next Frontier for Intelligent Transportation System

Dr. Qi Han, (Colorado School of Mines) Professor of Computer Science, Dr. Qichao Wang (NREL) Computational Transportation Scientist and Dr. Stanley Young (NREL) Advanced Transportation and Urban Scientist

The field of intelligent transportation systems (ITS) has long emphasized the use of advanced sensors to detect and react to a dynamic roadway environment. Although ITS technology has advanced significantly, even to the point of synthesizing optimal trajectories for vehicle paths, communicating the appropriate speed and path to the human driver remains problematic. Changeable message signs or variable speed limit signs provide directives to the human driver, but require drivers’ attentiveness for effectiveness, and limit direction to all drivers, not specific instructions to a specific vehicle/driver. Other attempts communicate directives through an in-vehicle interface, such as a display showing appropriate speed to arrive at a traffic signal during the green phase, or smartphone maps for navigating. The latter two require diversion of attention away from the physical environment to receive the directive of the cyber-physical system, which in turn creates an additional safety hazard. This Nexus proposal seeks to push ITS to the next level by using Augmented Reality (AR) to assist driving in complicated real-world environments. The AR enabled system will enhance a human’s ability to appropriately react to the dynamic physical environment, whether to avoid a potentially hazardous events (a vehicle failing to yield to a traffic light), or to appropriately govern the speed of the vehicle to seamlessly merge into traffic, or approach a traffic signal such that the arrival is synchronized with the onset of the green phase. Indeed, without AR, ITS technologies (i.e., active management of roadway traffic) will reach (and in many ways has already reached) fundamental limits of effectiveness. 

With AR, the system may provide humans with enhanced perception of the roadway environment (seeing around corners, knowing signal timing, coordinating movements with other vehicles) to gain significant efficiency and safety benefits. Such interactions could safely extend the driving abilities as people age, providing freedom and security of movement later in life as well as providing greater safety for vulnerable road users. Similarly, it could augment the perception of beginning drivers, helping to minimize driver perception errors typical of new drivers (such as clearance gaps).

 

Advancing Towards Resilient and Net-Zero Power Grids Through Connected Communities

Dr. Qiuhua Huang (Colorado School of Mines), Associate Professor of Electrical Engineering and Dr. Yuqi Zhou (NREL) Postdoctoral Researcher-Electrical Engineering, Power Systems Engineering

The rising number of natural disasters in recent years has consistently challenged the safe and reliable operations of power grids. Unlike transmission systems, which are interconnected in a robust, meshed network, the distribution systems (including microgrids) that directly supply power to residential customers are more vulnerable to system contingencies. As these residential loads are typically connected in a radial network topology, they are often at a greater risk of experiencing community-wide power outages and disruptions under overload conditions, equipment failures or severe weather events.

With the evolution of smart grid systems and advanced control technologies, the concept of connected communities becomes possible. The integrated networked system consists of multiple layers including power systems, transportation networks, and building management systems. The interconnectivity within the networked communities can facilitate resource sharing and improve grid resilience against grid emergency events. As electric vehicles, battery storage systems, renewables, and controllable loads become more widespread, effective coordination of these distributed energy resources is crucial for enhancing energy efficiency and attaining net-zero emissions. We plan to leverage our prior works and strengths, and develop emission-aware, resilient hierarchical control and optimization strategies for these connected communities while considering differential energy flexibility and social vulnerability within each community.

 

Scalable and Reliable Autonomous In-Field Heliostats Installation

Dr. Xiaoli Zhang (Colorado School of Mines) Associate Professor of Mechanical Engineering and Dr. Guandong Zhu (NREL) Senior Researcher and Group Manager, Center for Energy Conversion and Storage Systems

The topic of this university – national laboratory -industry collaborative proposal is “Scalable and Reliable Autonomous In-Field Heliostats Installation” with the objective to facilitate autonomous assembly and canting adjustments of commercial multi-facet heliostats directly in the field. This aims to minimize reliance on human labor while enhancing the precision of heliostats. The autonomous installation actions will be achieved through a mobile, multirobot system that cooperatively interacts with specially designed heliostats for safe and easy robot manipulations across the entire concentrating solar-thermal power (CSP) power plant. 

Installing heliostats and maintaining their accuracy consistently requires heavy human involvement. Although previous research has achieved measurement automation for CSP, such as Figure 1 Task overview of the autonomous in-field CSP installation Zhang, Xiaoli – #221 3 of 10 3 using unmanned aerial vehicles for sensing in NIO1 and UFACET2 developed at NREL, the heliostat installation process, including facet mounting and fine canting adjustment, still mostly rely on manual efforts. While robotics technologies have reached a state of readiness for elements of autonomous in-field installation, the unresolved issues of adopting and scaling up robotics installation with corresponding heliostat modification, optimizing robotic installation efficiency and maximizing power output are still open challenges. While robotics technologies have reached a state of readiness for elements of autonomous for in-field installation, there still exist several unaddressed challenges that prevent successful deployments of autonomous systems. These challenges include adopting and scaling up robotics installation with corresponding heliostat modification, optimizing robotic installation efficiency and maximizing power output. Addressing these challenges can help advance the automation capabilities for in-field installation and canting adjustment, which can ultimately reduce the construction cost to the target of $50/m2 by reducing the need for human labor. Further researching and understanding in the following aspects needs to be acquired in search for solutions for current challenges: 1) How to modify current heliostat structure to be manipulable by robots during installation and canting/recanting, 2) What features, such as robot’s configuration, resolution, lifting capability and control logics, are desired to autonomously install and calibrate heliostats, and 3) Understand the correlation between modification, cost, and power production for optimal power plant design.

 

Investigating the Solid Electrolyte Interphase in Sodium-Ion Battery Systems Using Mid-Infrared Microscopy

Dr. Yamuna Phal (Colorado School of Mines) Assistant Professor of Electrical Engineering and Dr. Lydia Meyer (NREL) Postdoctoral Researcher, Electrochemical Energy Storage Group

Sodium-ion batteries (Na-ion or SIBs) are emerging as promising alternatives to traditional lithium-ion batteries, offering scalable, cost-effective energy storage solutions. The recent significant funding opportunity announcement (FOA) from the DOE Vehicles Technologies Office (VTO) focusing on Na-ion battery technologies emphasizes the critical importance of innovation in this field. However, realizing the full potential of SIBs requires addressing key challenges, particularly regarding the role of the solid electrolyte interphase (SEI), and its contributions to cell performance and longevity. Spatially resolved chemical information of the SEI holds immense significance in elucidating the intricate mechanisms governing SIB performance and degradation. Traditionally, Fourier transform (FT)-IR-based spectrometers have been employed, but they suffer from limited spatial resolution, low signal-to-noise ratio and hence long acquisition times. The proposed concept of quantum cascade laser (QCL)-based infrared (IR) microscopy platform for monitoring the SEI aims to bridge this technological gap, offering scalable, real-time insights into SEI dynamics crucial for enhancing SIB performance and manufacturability. By utilizing IR microscopy to map the SEI, we aim to understand the nuanced effects of various candidate electrolytes on SEI formation dynamics, thus fostering the development of next-generation hard carbon anodes for SIBs with enhanced cyclic performance and cell stability.