Quantum computers work in a completely different way from conventional computers. Volkswagen demonstrates the practical use of this technology for the first time with a pilot project for traffic optimization in Lisbon.
Lisbon, the Portuguese capital on the Atlantic, is at the center of digitalization again this week. Tens of thousands of visitors are arriving from all over the world to attend the WebSummit, one of the world’s most important tech conferences. Of course, all these visitors aren’t restricting their movements to the exhibition grounds on the banks of the Tejo, but rather are traveling all over the city. In past years, this increased traffic has repeatedly resulted in the entirely analog problem of congestion.
This year, with the aid of quantum computers, Volkswagen will enable traffic to flow more smoothly in the streets of Lisbon during the WebSummit. To achieve this, nine MAN buses owned by the municipal transport company CARRIS have been equipped with a system developed by Volkswagen. Virtually in real time, the system calculates the quickest individual route for each of the participating buses and gives the drivers navigation recommendations by means of an app. This will enable the buses, which shuttle between the most important points in the city and have the highest passenger volume, to avoid congestion at an early stage and deliver thousands of people to their destination more quickly.
(Read more about the pilot project here)
The project has an important role-model function: in principle, it could be applied to any given city and any number of vehicles. Volkswagen is also using the project to demonstrate, for the first time in the world, a practical application possibility for quantum computers – a technology which experts have been discussing and researching theoretically for a long time, but whose practical implementation is still in its infancy.
How does a quantum computer work?
The way a quantum computer works is fundamentally different from today’s conventional computers. The memory of a traditional computer works with bits, the smallest information unit in information technology. Each bit can either take the value zero or one.
Quantum computers are designed differently. In their hardware architecture, they follow the rules of quantum mechanics and calculate with quantum bits (qubits for short). A qubit, and now the laws of quantum physics come into play, can be in the state of zero or one. But this state, however, can also be the result of a superimposition of various state combinations (known as superposition). This means that a qubit can have several states at the same time and store considerably more information than a single bit. A quantum computer can operate with all states at the same time in one calculation step. If several qubits are then coupled together, the number of possible combinations increases exponentially and so does the computing power.
Will there soon only be quantum computers?
No. Quantum chips are coprocessors which expand existing information technology. For tasks that “conventional” computers can easily solve, they will not be replaced by quantum computers. Quantum computers are ideally suited, however, to solve highly complex tasks far more quickly than traditional computers, or to solve tasks that conventional computers cannot solve at all.
In practice, today’s computers have the advantage that they work reliably and are robust machines. Quantum computers are technically extremely sensitive. Any external impact, including even the earth’s magnetic field, can disrupt the system. In magnetic spin qubits, the qubit architecture which is mainly used today, superconductors are used, i.e. materials that have no electric resistance. In order to function, however, they have to be cooled to extremely cold temperatures in the region of absolute zero. These apparatuses are correspondingly complex.
Working with quantum computers therefore requires specialist knowledge and methodological expertise, both of which Volkswagen Group IT is continually developing. In the traffic control project in Lisbon, both traditional computers and quantum computers are being used.
Which quantum computing systems exist?
Currently, the following two quantum computing architectures prevail, which are each suitable for different tasks:
An annealing system (annealing quantum computer) is designed to solve optimization problems. Simply put, an optimization problem deals with the question of how a resource such as time, money, or energy can be optimally used in a particular scenario. An annealing system finds the “sufficiently best” solution here, in other words the “energy minimum.” Volkswagen Group IT has implemented initial research projects for traffic control using an annealing system by the Canadian company D-Wave Systems.
A universal quantum computer (universal gate quantum computer) may be suitable for algorithmic problems of any complexity and nature. We say “may” because the algorithm development for practical application is also the subject of current research. The internet corporation Google, for example, is working on this technology. Specialists from Volkswagen Group IT are currently doing application-oriented research on a quantum computer by Google and testing its potential. Universal quantum computers are also suitable for traffic control.
How does Volkswagen use quantum computers in practice?
Martin Hofmann, Volkswagen Group CIO, says: “We at Volkswagen want to continue to expand our expertise in quantum computers and develop a deep understanding of how this technology can be used in a meaningful way in the business world. Optimizing traffic is one such possible use. Intelligent traffic guidance that uses the performance capacity of a quantum computer can provide meaningful support for cities and commuters.”
Quantum algorithms could be used for diverse objectives in this area. The Volkswagen specialists are developing both solutions for individual road users (single vehicles) and control possibilities for urban traffic planning. In addition to navigation tips to reduce travel time and avoid congestion, it would be conceivable to send vehicles information calculated in real time about available e-charging stations or free parking spaces. It is also imaginable to integrate overarching factors such as urban traffic management systems, public transport, or weather conditions.
What else are quantum computers helpful for?
Projects involving the simulation of chemical molecules and the interactions between their elementary particles in the quantum computer are already relatively concrete. Whereas traditional computers quickly reach their limits here, quantum computers promise a solution. “We can already do that today with simple molecules,” says Florian Neukart, Director of Advanced Technologies at Volkswagen Group IT in San Francisco. The result of this research could include previously unknown materials that are extremely resistant or particularly light and thus suitable for vehicle construction.
Another area in which quantum computers could help Volkswagen to make major advances is battery technology. Understanding what exactly happens in a battery at the atomic level could open the way to lighter and even more powerful batteries for electric cars. These would be findings that would flow directly into, for example, the recently launched production of battery cells in the Volkswagen Center of Excellence for Battery Research in Salzgitter. Every driver of an electric car would then have a concrete benefit from this research in their everyday life.
Publications by Volkswagen
- Florian Neukart, Gabriele Compostella, Christian Seidel, David Von Dollen (2019): System and method for predicting and maximizing traffic flow, US20190164418A1
- Pending: Florian Neukart, Michael Streif, David John Von Dollen, Tanja Graf, Thomas Schladt, Arne-Christian Voigt, Jonathan Edward Mueller (2019): Method for simulating electronic structure with quantum annealing devices
- Pending: Florian Neukart, Dyon Van Vreumingen, David John Von Dollen, Arne-Christian Voigt, Michael Hartmann, Carsten Othmer (2019): Method for finite elements-based design optimization with quantum annealing devices
- Van Vreumingen Dyon, Neukart Florian, David Von Dollen, Voigt Arne-Christian, Hartmann Michael, Othmer Carsten (2019): Quantum-assisted finite-element design optimization, arXiv
- Yarkoni Sheir, Leib Martin, Skolik Andrea, Streif Michael, Neukart Florian, Von Dollen David (2019): Volkswagen and quantum computing: an industrial perspective
- Gabor Thomas, Christian Seidel, Neukart Florian, Isabella Galter (2019): Assessing Solution Quality of 3SAT on a Quantum Annealing Platform, Springer
- Streif Michael, Neukart Florian, Leib Martin (2018): Solving Quantum Chemistry Problems with a D-Wave Quantum Annealer, Springer
- Neukart Florian (2018): Selbstfahrer: Wie künstliche Intelligenz das Auto der Zukunft beeinflusst, iX Magazin fuer professionelle Informationstechnik
- Neukart Florian, Seidel Christian (2018): A hybrid solution method for the Capacitated Vehicle Routing Problem using a quantum annealer, journal selection ongoing.
- Hsu Tin-Jui, Jin Fengping, Seidel Christian, Neukart Florian, de Raedt Hans, Michielsen Kristel (2018): Quantum annealing with anneal path control: application to 2-SAT problems with known energy landscapes, journal selection ongoing, arxiv: 1810.00194
- van Vreumingen Dyon, Neukart Florian, Von Dollen David, Bäck Thomas, Deutz André (2018): Quantum-assisted finite elements for design optimization, Frontiers in Physics, currently under review.
- Neukart Florian, Von Dollen David, Seidel Christian (2018): Quantum-assisted clustering, accepted in Front. Phys., arXiv:1803.02886
- Neukart Florian, Seidel Christian, Compostella Gabriele (2018): Verkehrsflussoptimierung mit einem Quantum Annealer, iX Magazin für professionelle Informationstechnik, 2, 2018
- Neukart Florian, Von Dollen David, Seidel Christian, Compostella Gabriele (2018): Quantum-enhanced reinforcement learning for finite-episode games with discrete state spaces, Front. Phys. 5:71. doi: 10.3389/fphy.2017.00071
- Goddard Phil, Mniszewski Susan, Neukart Florian, Pakin Scott, Reinhardt Steve (2017): How Will Early Quantum Computing Benefit Computational Methods?, SIAM News, 12, 2017
- Neukart Florian, Compostella Gabriele, Seidel Christian, Von Dollen David, Yarkoni Sheir, Parney Bob (2017): Traffic flow optimization using a quantum annealer, Frontiers in Information and Communication Technology, ICT 4:29. doi: 10.3389/fict.2017.00029
- Neukart Florian, Hofmann Martin, Bäck Thomas (2017): Artificial Intelligence and Data Science in the Automotive Industry, Data Science Blog and arXiv:1709.01989v1 [cs.AI]