There may be no one more excited about the role of computer science in energy than managing director at Duke Energy Initiative Kyle Bradbury.
“We have never seen such a major surge in the amount of data,” said Bradbury in a recent interview. “Soon, there will be very little that we can’t do with computer science.”
The Duke Energy Initiative works on fostering interdisciplinary collaboration for the development of cleaner energy systems. There, Bradbury works with computer scientists on data analytics and specializes in developing systems that learn with time.
Bradbury claims himself to be one of the few that are very optimistic about the future of energy and technology. But even he believes that considerable work in energy software is needed before it’s ready to save the planet. He sees this work being needed not only in the algorithms he designs, but also in the larger software industry.
“There is plenty of work to be done,” said Bradbury. “If you’re a computer scientist looking to develop for energy, you’ll have your hands full for a while.”
The list below outlines the four main ways that computer scientists can help fight climate change. It has been compiled using insights from energy systems and software researchers like Bradbury.
1. Develop Software for the Public Education of Energy Consumption
Bradbury is not the only one who sees room for improvement in energy related software. Dr. Ramesh Shankar is an energy researcher at EPIC, an organization focused on modernizing the grid. He believes informative software – software for informing the public — to be the biggest opportunity for computer scientists.
Shankar believes making electricity consumers aware of their responsibilities to be key in transitioning to a cleaner future. He also encourages smartphone application makers to take his ideas and develop educational apps for energy. In a recent interview, he mentioned the following:
- An app that shows typical household consumption of electricity on a monthly basis.
- An app dashboard for what the greenhouse gas emission impact is on a daily basis segmented by county, state, and nation
- An app that provides small actionable steps for electricity consumers and ways to track them.
- An app that shows the impact each individual can have if they use less energy or use more renewable energy.
Both Bradbury and Shankar expect a lot of change in the coming years. But unlike Bradbury, Shankar sees this change coming primarily from the consumers of electricity.
2. Develop More Robust Decision Making and Predictive Software Using Machine Learning
Software designed to identify trends in data to help make decisions can be classified as predictive and decision making software.
When asked about his thoughts on the current standing of such software, Bradbury laughed.
“Predictive decision making software may be the most scrutinized thing in energy today,” said Bradbury. “Virtually all aspects of it can be improved because of the complexity of the things it tries to model.”
He said that this same complexity is what makes creating the perfect predictive software practically impossible. Bradbury believes that pursuing such a goal is foolish and that developers will have better luck if they choose a more clever path.
“It’s impossible to account for all the variables in an electric grid, the climate, or any system for that matter, ” said Bradbury. “That’s why we need to create software that can evolve. We need software that can learn by itself.”
Thanks to massive data sets on the internet, he said machine learning – systems that learn automatically — will be absolutely essential to success.
“Researchers and engineers have done great work till now with predictive software,” said Bradbury. “But now we need more. More visualization software, better performance, and more use of energy data in helping systems learn.”
3. Help Develop More Customized Software for Collaborative Multi-disciplinary Research
Bryan Lampley, an engineer at Hoffman Building Technologies said, “Not a single one of us is smarter than every single one of us.”
Lampley believes that the fate of the energy sector lies in how well we can work together on big issues. He said that without insights from various disciplines, progress will be very slow – something, according to him we cannot afford.
“Climate modeling is a computationally intensive process that requires vast data sets and complex physical atmospheric models to accurately understand the factors that lead to changes in climate,” said Bradbury. “These require strong numerical programmers to work together with atmospheric scientists to produce these insights.”
Bradbury said that while current software for collaborative research is available, customized software is in high demand. He says that without a custom multi-disciplinary research platform, modeling something like climate would be next to impossible.
“It is just too difficult, and too complex,” said Bradbury. “Specialized software that takes into account what the researchers want will help make the process far more efficient.”
He said that this is true for all collaborative research in the field of energy. Bradbury mentioned collaboration between computer science and economics as another example.
“These predictive and prescriptive models will require detail-oriented programmers to work with people knowledge of economics and policy to help produce accurate decision-making tools,” said Bradbury. “Good software for research will allow them to do this efficiently.”
4. Help Improve Cybersecurity in Large Company Networks
Lampley said that lack of proper cybersecurity is huge in preventing large, connected networks from taking off. He believes that the risk for hacking is too great right now.
“Cybersecurity is 100% on the forefront of all of these companies’ minds” said Lampley when asked about the potential for hacking in large company networks. “They definitely don’t want all of their private sensor data exposed to the internet. Would you?”
He said someone could hack in and cut all the power off for an energy company.
“This is absolutely something that simply has to be addressed.” said Lampley. “We have got to have some security when it comes to large connected networks.”
Keith Dickerson, manager of smart infrastructure at Schneider Electric agrees and said “this is a very contested topic and we need more computer scientists working on this issue.” He says that without it, data open to the public would be far more harmful than useful because of hackers with malicious intent.