Humanity is facing unprecedented challenges that will require the full engagement of research universities. This chapter identifies grand challenges, emerging topics, and requisite methods that will be addressed at research universities over the next 10 to 25 years. We have organized these research topics into the following three categories, recognizing that they intersect and overlap:
Societal Challenges: These include food, water, sanitation, and hygiene; healthcare and medicine; transportation, energy, and infrastructure; cybersecurity and privacy; diversity, equity, and inclusion; education; and security, defense, and prevention of nuclear terror threats.
Innovations: These are crosscutting and broadly enabling in nature, and include advanced materials; artificial intelligence and machine learning; data science and computer systems; and quantum systems.
Discovery: A variety of research topics are not necessarily pursued for some applied end, but for reasons of curiosity, understanding, enjoyment, contentment, and beauty.
The overarching thesis of this chapter is that university-based research can contribute to improved living conditions and quality of life, by contributing solutions to food, water, sanitation, energy, health, well-being, climate change, community planning, infrastructure, security, and education [Valero 2019; Wowk 2017]. University research can also develop underlying technologies, tools, and methods of analysis that can lead to innovative future solutions, such as artificial intelligence or advanced materials.
3.2 Societal Challenges
Food, Water, Sanitation, and Hygiene
Sustainably supplying food, water, and energy to all will be increasingly challenging because of population growth, increasing standards of living, and climate change. Innovation will be crucial in augmenting supplies, improving distribution, reducing waste, increasing efficiency, and reducing demand. In addition to pointwise innovation and technologies, holistic, systems-oriented approaches are required.
Food. Research and education are needed to produce more food with fewer inputs, including water, cereal grains, pharmaceuticals, and labor. Advances in agricultural technologies, data collection, and computational science provide opportunities to further enhance efficiencies and increase yields. For example, sensors can detect and diagnose plant diseases to reduce lost agricultural productivity [Mahlein 2016]. Precision applications of pesticides, herbicides, and fertilizer can dramatically reduce agrochemical use without compromising yields [Schumann 2010]. A better understanding of the microbiome in agriculture will improve soil structure, increase feed efficiency and nutrient availability, and boost resilience to stress and disease [NASEM 2018]. Selective breeding, genetic engineering, and gene editing could be used to develop crop varieties that maintain productivity under changing climate conditions [NASEM 2016]. Advances in low-cost sensors and communication tools could provide guidance to farmers on appropriate application rates of seeds, water, and fertilizer to maximize yields and prevent unnecessary inputs.
Technologies and systems along the entire food chain — including harvest, transportation, processing, and storage — are needed to reduce food loss from farm to plate. Protective films can lengthen shelf life, possibly without refrigeration [Sharma 2017]. Low-cost sensors that indicate food quality and safety could further reduce food loss.
Water, Sanitation, and Hygiene. Safe drinking water, sanitation, and hygiene are paramount to improving standards of living and quality of life [Hutton 2017]. Research can lead to improved water treatment, handling, and storage; improved toilets; on-site excreta management; septage and sewerage management; and menstrual hygiene management.
Wastewater reuse is more expensive than conventional water supply alternatives such as imported water and groundwater, and public acceptance of potable reuse remains a challenge. Advances are needed to reduce the cost and energy requirements of alternative supply treatment and to develop sensors and mitigation approaches for contaminants [CDC 2020], such as membranes that remove specific pollutants while allowing nutrients to pass through. Technologies that improve the recycling of wastewater and sewage treatment so that water can be used for irrigation or industrial purposes are still needed. Recycled water has the potential to resupply aquifers, but effective purification methods and rigorous safeguards are necessary.
Desalination is currently costly and requires large scales. Nanotechnologies such as nano-osmosis using nanotubes, nanofibers, and heterogeneous catalysts have exceptional potential for their filtering abilities. In addition, technologies are needed to reduce water use. Agricultural irrigation consumes enormous quantities of water. For example, in developing countries, irrigation often accounts for more than 80% of total water use. Improved technologies can reduce agricultural water demand while providing enough water to cleanse the soil. Work is also needed to reduce water loss in urban supply systems.
Healthcare and Medicine
Healthcare and medicine are heavily driven by research. For universities, research related to the causes of cancer, diabetes, neurological diseases, viral transmission, and respiratory illnesses and, ultimately, mitigating symptoms, can improve the quality of life and life expectancies [NAM 2017]. In addition, there is continued opportunity for advancement of medical technology and health service research, which includes brain research, application of artificial intelligence in the identification of symptoms and to look for abnormalities in X-rays or CT scans, and systemwide approaches that incorporate social determinants to health outcomes [NAM 2018]. Research is needed to improve the force-feedback mechanism in robotics, to allow for remote surgical operations where a remote surgeon can perform surgery on an individual without the need to travel. Research into effective approaches to data integration, strategies for patient-centered care, health disparities, and the embedding of research into the care environment are also important areas [Bestsennyy 2020]. For example, through the Covid-19 pandemic, health professionals significantly increased the use of telehealth technologies to prevent potential patient and practitioner exposure. Research is needed in developing novel data collection strategies through personal wearable devices.
Coupled with these new technologies is the need for more secure data transmission and record keeping. The integration of volumes of data generated for each individual along with techniques tailored for big-data analytics may enable us to achieve precision medicine [Cattell 2013].
Finally, there is overwhelming evidence of health inequities in the United States [NASEM 2017]. Research is similarly needed to identify and understand these inequities and develop solutions that are frugal, robust, and easily accessible.
Transportation, Energy, and Infrastructure
Transportation is fundamentally important to humanity and integral to urban development, mobility, and economic growth. The primary modes of interdependent transport include road, rail, aviation, maritime, and pipeline. Opportunities for research include developing new, efficient strategies for manufacture, design, construction, operation, inspection, maintenance, and renewal or disposal of associated systems. Linking scales of research to practice via knowledge sharing, standardization, and technological adoption remains a particularly important challenge. These issues can be addressed through the development of sustainable materials, robust sensors, integrated GPS and wireless communications, battery power storage, automation, smart systems (for damage detection, automatic adaptation, and damage tolerance), and policy research [Polzin 2016; Polzin 2018]. For air travel, research is needed in the development of electrified airplanes, as well as supersonic and hypersonic flight [NASA 2020].
Energy is another area where the need for sustainability will drive major changes, and research will play an important role. Increasing the efficiency, decreasing the cost, and increasing the life of solar devices is just one example of vital research areas. All elements associated with the use of electric power — power electronics, wide bandgap materials, more efficient electric machines (e.g., motors), smart grids, and decentralized control — will also become increasingly important. Work is also needed to reduce the cost of net zero carbon chemical energy carriers, such as hydrogen or renewable hydrocarbons. Research is needed to reduce costs and increase reliability of decarbonized power generation [DOE 2015]. Finally, research is needed in energy efficiency and to reduce energy needs, such as in buildings, manufacturing, and transport.
Many of the improvements discussed here can only be delivered in a cost-effective way through public infrastructure. Aging infrastructure systems pose considerable risks related to maintenance, service, and upgrading. Future research areas include technologies rooted in the Internet of Things (IoT) designed to enhance condition monitoring, response prediction, maintenance, and operation. The grand challenge is to develop smart and adaptive materials (e.g., self-healing, self-sensing, self-cleaning, etc.), responsive sensors, data systems, and wireless technologies that provide interactive communications. Moreover, conception, development, and funding of infrastructure projects, in both developed and developing countries, will benefit from social science research.
Cybersecurity and Privacy
From power grids and air transportation to healthcare systems, physical infrastructures are increasingly controlled by computer systems, creating opportunities for adversaries to steal sensitive information, disrupt operations, or even destroy critical assets, all carried out through cyberspace. We have already seen a significant increase in cyberterrorism in financial systems and the power grid. Moreover, social infrastructure connections pose risks to fair governance and democracy, elections processes, and the safety of people. Research is needed to develop approaches to protect critical cyber-physical infrastructures and encourage partnerships among security and domain researchers, as the detection, mitigation, understanding, and attribution of attacks necessitates domain-specific expertise. There is also a pressing need to understand how people interact with their computers and information culture and to protect personal data and preserve privacy.
Diversity, Equity, and Inclusion
While equity and inclusion are values that drive the subjects and methodologies of our research, understanding and quantifying the many ways in which technologies, social structures, and a host of other factors create or reinforce inequity is a research topic in its own right. This work involves studying and supporting strategies focused on racial, gender, place-based, and other forms of equity, and identifying the structural barriers facing various communities. It is vital that researchers collect data, quantify change, and identify correlations to identify inequity. Research is also needed to identify the biases and discrimination that flow from new technologies, and to suggest approaches to harness their benefits while mitigating potential harm. For instance, while developing artificial intelligence (AI), we must make sure that our data sets do not implicitly bias a technology. Research is needed to evaluate disparate impacts that new technologies might have on workers. Finally, research is needed to understand bias in organizational decision-making, and how to prevent it.
Higher education has dramatically evolved in the past decade, incorporating new ideas about classrooms, curricula, pedagogies, learner profiles, and the overall system, aimed at delivering more effective learning. Research will be foundational to devising pedagogical approaches and developing tools for building curricula that do not just pique learner interests and abilities, but also instill critical, creative, and cognitive thinking, acting as the base for lifelong skill development. Evidence-based educational ideas and platforms will integrate both previous and real-time data in conjunction with machine learning to inform practice in a timely fashion, to improve student outcomes, and to address current challenges. Data integration and data mining on a national scale will also provide larger sample sets that would prevent the ambiguities present in smaller data sets.
Security, Defense, and Prevention of Nuclear Terror Threats
National security not only includes guarding against military threats, but also fortifying a country’s security across its economy, healthcare system, industries, supply chains, technologies, energy resources, and infrastructure. All of these involve a host of research problems, such as analysis of supply chains or geopolitical influence networks, or detection techniques of nuclear materials.
Materials are an integral part of all physical technologies that cut across essentially every application field. As a result, designing and creating new materials with enhanced properties can have a transformative effect on the performance of next-generation systems, devices, components, and applications. In the coming decades, we can expect that materials will be developed that will have application-specific, performance-driven attributes.
Developing targeted materials requires the ability to model, fabricate, and measure properties across diverse length scales. The goal is to develop methodologies that enable stretching the limits of performance of existing materials, or designing and creating new materials for specific applications. These techniques can be used to make metals lighter and stronger, ceramics tougher, polymers degradable, and composites more affordable. Several grand challenge problems require advances in materials: new materials can significantly improve the performance of energy storage technologies, including batteries, capacitors, superconducting magnets, and flywheels [NASEM 2019]. New materials can make dramatic improvements in separation and filtration problems, such as filtering clean water or carbon capture. More efficient and sustainable construction materials could dramatically improve the energy efficiency of buildings and reduce the environmental impact of urban infrastructure.
Artificial Intelligence and Machine Learning
As an enabling technology, the recent and projected growth of AI has the potential to make the next century one of the most transformative periods in human development. While the recent advances of automatic speech recognition, image recognition, and self-driving vehicles have been impressive, we will see further advances in healthcare, education, security, manufacturing, and scientific discovery. Already able to identify and locate some forms of cancer that are too subtle for the human eye to detect, AI systems in the future will have the potential for personalized, round-the-clock monitoring and diagnostic capability through low-cost and minimally invasive digital assistants. At the global scale, AI algorithms will make spatiotemporal connections between reported symptoms of patients and social media to identify, locate, and track potential communicable disease. Further, future AI systems will provide cognitive support to an aging population as well as early-warning detection and personalized mitigation strategies for depressive episodes in susceptible patients.
AI can improve predictive maintenance of airplanes, ships, bridges, and other complex systems and infrastructure, streamlining operations and ensuring a robust supply chain. AI can also improve efficiency through human-machine interfaces or human performance enhancements, whether physical with exoskeletons or mental with decision support systems. Autonomous systems can also remove the human from dangerous or tedious tasks, such as in nuclear power plants or war zones. Further, AI has the power to improve education, starting at the early stages of childhood development and continuing into retirement, such as through aiding teachers in designing personalized lesson plans to handle the different developmental speeds of pupils.
While artificial intelligence capabilities enable human-like perception and processing across the technical spectrum, the novel attributes of AI couple into control processes to create new vulnerabilities and exacerbate existing security challenges. Future AI systems will need to be able to identify deepfakes, ensure both national and personal privacy, and be robust against poisoning through manipulated data sets. Research is needed to integrate domain-specific scientific principles, expert feedback, and uncertainties in the algorithms to enhance the capability, accuracy, and defensibility of these models. One opportunity for domain-specific machine-learning research is in constructing modeling languages and frameworks that facilitate the inclusion of domain knowledge into training.
Another crucial need in scientific machine learning is to maximize the amount of learning in situations with scarce data sets, and to actively choose data or new experiments wisely. In active learning, there are opportunities for research to improve the design optimization process, especially for systems that are dynamically evolving in time. In particular, the availability of a physical model can provide predictions that enable look-ahead for future design choices. Research is needed to enhance these models to ensure the methods are stable, robust, and bias-free. To meet these challenges, an AI-friendly workforce is required, not as a separate field of study, but integrated throughout STEM.
Data Science and Computer Systems
Over the past decade, the ability to rapidly generate, store, and analyze data has had a significant impact on almost every sector. Data is now pervasive in nearly all aspects of modern life. Global internet traffic is greater than 200 Exabytes (1018 bytes) per month. There are billions of devices connected to the Internet of Things (IoT), generating massive amounts of data. Additional data is collected for research, medical, and national security applications. In order to be useful, all this data must be securely acquired, distilled, labeled, aggregated, and stored in an accessible way. As our ability to generate data continues to grow, we must continue to develop technologies for low-cost data storage and high-performance computing for data analysis, as well as efficient algorithms and optimization techniques. We need advances in edge computing systems to reduce communication costs and improve efficiency. End systems need to operate in real time and handle the deluge of data by providing real-time tagging and aggregation. Further, as devices fail or become compromised, techniques for identifying and mitigating malfunctioning sensors must be developed to provide robustness to an increasingly networked world. Finally, gains in advancing the data fabric must preserve privacy, and research should specifically address robust data science methods.
Proliferation of data will have an enormous impact on all sectors of society, including industry, education, medicine, science, national security, and government. Big data sets abound in genomics, systems biology, and proteomics. Advances in electronic medical records, computational phenotyping, personalized genomics, and precision medicine are driving predictive, preventive, and personalized healthcare. Large-scale data sets providing a microscopic view of materials, and scalable modeling and simulation technologies, are paving the way for the accelerated development of new materials. Advances in sensors and Internet of Things technology enable energy infrastructure monitoring. Data analytics brings unparalleled efficiencies to energy production, transmission, distribution, and utilization [Rockefeller 2020; Lytras 2019].
Dramatic improvements in computer hardware have driven data science advances. Hardware improvements deliver performance that unlocks new capabilities. Novel computing systems will be needed to leverage both new technologies and new architectures. While there has always been a strong link between the availability of hardware and advances in software and applications, the rapid growth of transistors per chip allowed hardware and software to be developed relatively independently. In the future, this will change as novel computing systems are developed to optimize algorithm performance for domain-specific applications. Conversely, the discovery and implementation of novel hardware technologies will drive innovation in new types of algorithms and applications.
There are several potential approaches to building novel computing systems, some more disruptive to the current development cycle than others. Some novel computing systems will not only use new technologies for the logic elements but will also use radically new computing paradigms. Potential applications include the design of complex electronic and optical materials, the integration of molecular-scale circuits for biological applications, and therapeutics that have programmable control over drug delivery [Passian 2019]. DNA and other synthetic molecules can also be used to store archival data with significantly reduced size or power requirements. Potentially among the most disruptive forms of novel computing systems are quantum computers, which have the potential to surpass the scaling limits of classical computers for certain algorithms, discussed further in the next section.
Quantum computers could have a transformative effect on society [Möller 2017], by enabling better solutions to a class of large combinatorial problems than any known classical solver. This means more efficient logistics and supply chain designs, better circuit designs, and superior protection against malware attacks, among other applications. Similarly, quantum materials are a promising and broad class of materials that should enable technologies of the future, just as advanced materials enable technologies like MRIs, biosensors, and disk drives today. But quantum physics also sets the limits of what is possible in other types of systems, and research groups around the world have demonstrated lab-scale devices that harness entanglement and/or coherence that outperform classical solutions. The transition from the laboratory to operational devices faces many scientific and engineering challenges, but quantum systems will likely play an important role in next-generation devices across a wide spectrum of applications. Bridging the gap between the current ability to characterize large quantum systems and the capacities of the smallest workable quantum computers is the immediate grand challenge for the field.
Quantum sensors are positioned to deliver solutions to critical problems, but techniques must be developed to extract this sensitive signal from the background, and technologies must be developed to decrease their size and power requirements.
Atomic magnetometers have the potential to transform a number of important fields. They could be used to map brain and heart activity without cumbersome cryogenic cooling, which would contribute to our ability to diagnose injury and disease and to understand the way the brain works. They can be used for magnetic anomaly detection, potentially allowing for the early detection of UAVs, submarines, and underground buildings and structures — all applications with enormous implications for national security. They could also be used for wide-area aerial high-resolution magnetic surveys to detect archeological or geological structures, mineral deposits, contaminated soil or maritime sediment, unexploded ordnance, or abandoned vehicles. The same technology can be used to do navigation using precise magnetic maps of the earth’s crustal field.
Atomic clocks currently provide primary time and frequency standards for the United States as well as the very precise time data used for GPS signals. Efforts are underway to decrease the size, weight, and power of these systems while maintaining a high level of precision and stability. Improved compact atomic clocks could allow for long periods of silent navigation in environments where GPS is not available — underwater, underground, or in the battlefield where GPS has been jammed. Distributed systems of precise atomic clocks can lead to coordination of signals from multiple sources that can be used for improved geolocation, image analysis, and target tracking.
Antennas could surpass classical limits and allow the construction of compact communications systems. Measurements of quantum noise can enable the next generation of world-class instruments and facilities to create fundamental scientific knowledge and probe the cosmos.
Curiosity and Understanding
Basic or discovery research is focused on the acquisition of general knowledge, understanding nature, and predicting outcomes or phenomena. Humankind has always sought to expand horizons and explore new frontiers. Such endeavors are rooted in curiosity about the unknown or to understand what is being observed. These topics include pure mathematics, the search for life on other planets, and understanding the chemical origins of life.
Such research is often performed without thought of the direct applications or practical ends. There is often a 10- to 20-year gap between fundamental discoveries and subsequent application [MIT 2015]. Moreover, intentional support of fundamental research contributes to the broader university culture that values the development of thought, the growth of ideas, and investment in intellectual talent.
Enjoyment, Contentment, and Beauty
Research in the creative arts represents an important area for any major research university. Art research offers the proposition of new concepts for the problems of contemporary society and the ability to offer alternative solutions or propose new questions, based on non-exclusive rational modes of thought. Art research fosters both conceptual and non-conceptual thinking. It spans multiple disciplines, including design, music, and composition, and offers other unique perspectives. For example, elements of music theory can be taught using mathematics. Architecture and building models can be directly influenced by artistic design elements and formulas. Art research also has a broader societal impact in existential enhancement (e.g., enjoyment, contentment, and beauty) and leisure activities as well as the economic impact of tourism and cultural industries.