Springe direkt zu Inhalt

Into the Future with Quantum Bits

Quantum computers mark a revolution in artificial intelligence.

Feb 21, 2020

On track for the future. Deutsche Bahn AG plans to use quantum-assisted algorithms to optimize train schedules.

On track for the future. Deutsche Bahn AG plans to use quantum-assisted algorithms to optimize train schedules.
Image Credit: Picture Alliance - Walter G. Allgöwer

Quantum computers are still in their infancy, but they could play a huge role in shaping society just a few years from now. Unlike conventional computers, quantum computers do not work on the basis of the laws of classical physics. The way they perform calculations follows the rules of quantum mechanics, and those rules alone – which makes them many times faster. “They can perform calculations in seconds that would either be impossible for a traditional computer or take infinitely long,” says Adrian Paschke. “This means they supply the computing power needed for the next step in the development of artificial intelligence.”

Paschke is a professor at the Institute of Computer Science at Freie Universität Berlin and also serves as the director of the Data Analytics Center at the Fraunhofer Institute for Open Communication Systems (FOKUS). He is currently involved in developing a platform for quantum artificial intelligence (QAI), which is being created as part of a consortium between Freie Universität, Fraunhofer FOKUS, and StoneOne AG. The “PlanQK” project, which is receiving funding from the German Federal Ministry for Economic Affairs and Energy, aims to advance the development of solutions in quantum artificial intelligence (QAI) for industry by developing a knowledge platform.

“Today, QAI is on the cusp of initial applications in industry,” says Paschke. “But using it is still difficult, since the technology is very expensive and the knowledge is spread across many private and public research projects.” This results in high barriers to entry, he says, especially for small and medium-sized companies. “Our goal now is to create a central pool of knowledge that is accessible for development and application purposes,” Paschke says. “We are also building a system that will allow specialists and developers to network with users, customers, and advisors.” There are 15 partners from the private business sector participating in the project. Use cases are being developed in cooperation with these partners. The platform will then show how the problem in question could be solved with a traditional AI algorithm and what quantum-supported alternatives are available. “We will select the most suitable algorithms from the research literature and then test which hardware is suitable for the algorithm and how the algorithm behaves,” Paschke explains.

One of the project partners is Deutsche Bahn, which hopes to use the platform to tap into new possibilities for optimizing train schedules. A financial sector company that specializes in fraud detection is also participating, testing algorithms that automatically search finance flows for irregularities and unusual patterns and can thus help authorities fight financial crime. Another project involves the development of an algorithm to predict the likelihood of pipe breaks in buildings.

“Quantum computers are so fast because they use quantum bits, or qubits, to perform calculations,” Paschke says. While normal bits, the computing units used by traditional computers, can only assume two positions – 0 or 1 – and only one of them at a time, qubits have what is known as a superposition state. “They can assume an intermediate state from both positions at the same time for a certain time span,” explains Paschke. At the same time, qubits are interconnected, with the state of one qubit directly affecting those around it.

“A traditional computer can only perform one calculation after another,” Paschke says. “A quantum computer, on the other hand, performs a large number of calculations at the same time.” In addition to the open knowledge platform, on which the project’s findings will be freely accessible, the PlanQK project also plans to open a “marketplace” where companies can sell proven applications, like in an app store, for further use. “We are working with a number of companies that have their own quantum computers for this,” says Paschke. The programs could then be executed via the platform on these companies’ computers, so future customers would not have to have a quantum computer of their own. And the next step toward the future is clear to computer science professor Paschke as well: development of a quantum computer in Germany.

This text originally appeared in German on February 15, 2020, in the Tagesspiegel newspaper supplement published by Freie Universität.