IE 11 Not Supported

For optimal browsing, we recommend Chrome, Firefox or Safari browsers.

In the Lab: Researchers Work on What's Next in Tech

The latest and greatest technologies often start as projects in university laboratories across the country. Here are a handful of innovations in the works that could transform gov tech in the coming years.

orange-lit beakers as part of a chemistry set on a black background
Government has problems: ensuring citizen voices are heard, keeping first responders out of harm’s way, mining agency data to support policy decisions without compromising privacy, etc.

The solutions to these and other challenges are emerging in university research labs around the nation. From robotics and cryptography to artificial intelligence and machine learning, academia is on the cutting edge when it comes to developing technologies that could someday impact the work of state and local governments — and the world.


At Rice University, synthetic organic chemist James Tour is looking to alleviate the crushing burden of plastic waste, while simultaneously reducing harmful emissions that are fueling climate change.

“Every state and local government is interested in reducing CO2 emissions. And every state and local government is beset with massive amounts of plastic waste,” he said. Nano-scale chemistry might make it possible to address both problems with a single solution.

Tour is experimenting with the idea of upcycling high-density polyethylene, one of the most abundant types of plastic waste. He’s looking to break down the material and reconfigure it into low-cost filters that could be inserted into the flue stacks of power plants, thus trapping carbon emissions before they can get loose in the atmosphere.

Rice University graduate student Paul Savas feeds raw plastic into a crusher. Researchers want to use plastic waste to sequester carbon dioxide from flue gas.
Rice University graduate student Paul Savas feeds raw plastic into a crusher. Researchers want to use plastic waste to sequester carbon dioxide from flue gas.
Jeff Fitlow/Rice University
To create the filtering materials, Tour’s lab first turns plastic into powder, which then is mixed with potassium acetate and heated at 1,112 degrees Fahrenheit for 45 minutes. “When we convert the plastic waste into this carbon material, with nanometer-size pores, it traps carbon dioxide coming out of a flue gas,” he said.

Once the carbon dioxide is trapped, “you can do whatever you want to with it: Pump it down a hole and it stays there for thousands of years, or convert it into useful products. Either way, we trap it, rather than using our atmosphere as a waste dump,” he said.

(Tour’s process also produces a wax byproduct, which can be recycled into detergents or lubricants.)

Government would likely work in cooperation with private industry to bring such a solution to life. Once the technology is commercially mature — Tour says that’s still about five years off — governments could, for example, require its use in power plants as part of the regulatory structure, and could make plastic waste readily available for such reuse as a component of an overall waste management plan.

“Government has an interest in reducing carbon dioxide emissions, but you can’t just snap your fingers and say, ‘Carbon dioxide be gone,’” Tour said.

By diverting plastic waste in support of that effort, local authorities could solve two civic problems at once: keep the stuff out of the landfills, while simultaneously addressing climate change.
A training session for facilitators learning to use the Local Voices Network at MIT.
A training session for facilitators learning to use the Local Voices Network at MIT.
Deb Roy/MIT


At the Massachusetts Institute of Technology, Deb Roy wants to ensure that resident voices are heard. As director of the MIT Center for Constructive Communication (MIT CCC) and a professor of Media Arts and Sciences, he’s leveraging artificial intelligence in support of a technology known as a Local Voices Network, or LVN.

“Through a powerful combination of AI and human listening, LVN enables organizations to gather people for recorded small group conversations around their life experiences, to make sense of the conversations they collect, and to surface community voices into public dialog and decision-making,” said Roy, who is already operationalizing the technology in his role as co-founder and chair of the nonprofit Cortico.

LVN combines human storytelling and machine listening, with the AI capabilities packaged in a platform that supports audio capture and uploading, management of conversation archives, semantic content analysis and visualization, and online posting of audio highlights.

Researchers are working “to develop and integrate even more powerful capabilities — such as conversation analytics, and novel ways of enabling data access and ownership — into the LVN platform,” Roy said.

The public sector could look to such technologies to elevate citizen voices.

“As we listen to people in recorded small group conversations across the country, we hear one sentiment again and again: ‘When we speak up, we don’t feel heard,’” Roy said. In-person forums such as town halls and open meetings also fall short: They typically attract the “usual voices” of the same committed community members, “in what are often symbolic, ineffective efforts to capture real community input.”

Surveys and focus groups likewise rarely capture the rich breadth of people’s lived experiences. “It’s clear that we need better ways of listening to the experiences of real people in order to spark decisions that are informed by community members themselves,” he said.

He describes LVN technology as a potential means to amplify people’s life experiences in support of a stronger democracy. This can happen in two directions: “laterally,” by improving listening and learning across the divides that isolate people from one another; and “vertically,” by informing the policy- and decision-making with community input grounded in lived experience and inclusive of often-underheard voices.

In the effort to promote LVN, he encourages government agencies “to build transparent community input practices at scale, making it possible to efficiently and meaningfully include a range of perspectives, ideas and needs in decision-making processes,” he said. “In this way, we imagine fostering communities of individuals who feel heard by their governments and, as a result, are more empowered to participate in civic processes.”

Some are already putting this into play. Madison, Wis., used the technology in its police chief selection process, and Cambridge, Mass., put it to work in the process of hiring a city manager. The New York City Department of Health and Mental Hygiene used it to collect stories to inform a citywide wellness road map, and Chicago officials are documenting conversations as part of We Will Chicago, a three-year citywide planning effort.


Richard Voyles is looking to robotics as a means to elevate a range of civic functions. “I’m working on a number of projects, from agriculture to manufacturing to emergency response and even medical robotics,” said Voyles, the Daniel C. Lewis Professor of the Polytechnic at Purdue University.

It's more than just the robots themselves.
He’s not talking about robotics in the conventional sense (e.g., Robby the Robot). In the academic world, “robotics is anything that brings together sensing, computation, and then acting on those,” he said.

For example, consider cow digestion. What goes on inside a cow is of great civic interest in a world where we’re looking to feed an ever-increasing population. To that end, Voyles is experimenting with “ingestible robots.”

These tiny devices “operate inside the cows to monitor their health and well-being,” he said. “We can make more efficient animal products in a world that relies an awful lot on animal protein. There is a growing need for food in the world, and this is something that all governments are concerned about.”

Another effort focuses on what Voyles calls “programming by demonstration,” a robotic-driven means of sharing information.

“When a parent teaches their child to tie their shoes, they don’t write a procedure down in a foreign language. They just show their kids what to do and they help them learn by example,” he said.

“That’s what we’re doing: Someone who knows how to assemble a fuel injection unit or some other piece of equipment can show the robot what to do, and it can learn from the demonstration,” he said. This could make robotic processes more readily available to state and local leaders looking to leverage such capabilities.

But government may be even more interested in the work Voyles is doing in the area of emergency response.

In a natural disaster such as a hurricane, tornado or earthquakes, “emergency response robots are really important for going where humans can’t go, doing things that humans often can’t do,” he said. “With growing climate instability and worsening weather, we’re seeing more and more need for robots that can help humans look for survivors, to find people who might be hurt but are still alive.”

Rescue work is inherently dangerous, and the need to keep first responders safe can often slow down an operation. “If an emergency responder gets trapped in rubble, gets caught in a secondary collapse or something like that, typically it’s going to take 10 other people 10 hours to get that person out,” he said. “You’ve lost 100 person-hours of rescue time.”

Robotic solutions promise to make such operations safer and ultimately more effective. To move along this path, Voyles said, civic entities will need to do some upfront planning.

“It’s more than just the robots themselves,” he said. “When you’re talking about robots for things like this, you also need to have communication [with those robots], and of course in a wide-scale disaster like a hurricane or an earthquake, often the communication has been wiped out.” To make effective use of these emerging tools, planners thus will need to incorporate emergency communications into the response protocols.

There’s a potential added bonus here: By planning for robotic deployments, emergency managers will have an opportunity to rethink their overall strategies.

“If we can plan motions for robots, we can also plan more effective evacuation routes for cars and people. We can plan evacuation strategies for stadiums and arenas,” Voyles said. “We can apply the software we use for robots to simulated humans. There are all these potential spillover effects.”
Robotic police in Dubai can help residents with tasks like paying traffic fines.
Robotic police in Dubai can help residents with tasks like paying traffic fines.


An assistant professor in the department of computer science at the University at Albany, State University of New York, Chinwe Ekenna also has her eye on robots. She’s leveraging AI to make them operate more effectively, which in turn could have repercussions for government.

Her work looks specifically at robotic motion-planning algorithms.

“Mathematical topology variants are being introduced to improve the usability of previously complex algorithms and scalability across different robot platforms. These algorithms help improve informed methodologies that exploit various AI and machine learning approaches,” she said.

We need this, she said, because faulty algorithms can get in the way of robotic operations, while improved algorithms can improve the outcomes significantly.

“When it comes to human-robot interactions, appropriate feedback on failures can aid in improving diagnosis as well as increasing robot safety,” she said. With better algorithms, operators can predict obstacles in a robot’s path, for example. “Failures are more accurately recorded, and more plausible paths are developed in the process.”

All this matters at a time when governments are looking to robotic processes in support of citizen service. In Dubai, for instance, the public interacts with robotic police assistants to report crimes and pay traffic fines. Other researchers point to the potential use of robots in everything from tax processing to supportive roles in medical care.

“Service robots [could] assist human beings in certain redundant, dirty tasks,” Ekenna said. Street and highway cleaning, lawn mowing, adaptive traffic control — smarter robots could help with all of these.


A professor of computer science at Boston University, and the director of the Center for Reliable Information Systems and Cyber Security, Ran Canetti says advanced cryptography could be a boon to government entities.

“Cryptography has legs in mathematics, in logic, in computer science, in algorithms — and it is smack in the center of society, of human experience with computers. It is in some sense the science of finding mathematical solutions to social problems,” he said.

What can it do? Take, for instance, a law enforcement application.

“You have people looking at child sexual abuse material, which is illegal to own and sell. Law enforcement is trying to catch them and bring them to trial, and they’re using hidden software tools, so that people will not see it and go around it,” Canetti said.

“When people get caught and are brought to trial, defendants have the right to inspect the evidence against them — to know how it was collected and to verify any issues with the software,” he said. “At the same time, there is clear interest for the government, for law enforcement, for the public in general, to keep this software secret in order to keep it effective.”

Cryptography offers a way to cut that knot. “We have actually been building a prototype of a system that can get around this. If law enforcement builds their tools in a certain way, using cryptography, they can actually prove out the different properties of the software without revealing the secrets of the software,” he said.

Government could apply the same strategies in areas such as trade secret litigation. With the right cryptography, “I can show that you used my secret algorithm, without actually exposing my secret algorithm,” he said.

An even more urgent use case has to do with the government’s role as an aggregator of civic data. State and local entities need to leverage often-sensitive personal data for statistical purposes, in order to shape policy. But they need to do it in a way that protects privacy.

“How can we share that kind of information selectively? This looks like a lawsuit waiting to happen,” he said. But cryptography makes it possible. “It allows you to compute joint statistics on the secret information, while keeping the actual information hidden.”