“Quantum Computing Optimization Challenge” with industry partners starts – QAR-Lab of LMU Munich pushes applications forward
"Quantum Computing Optimization Challenge" with industry partners starts - QAR-Lab of LMU Munich pushes applications forward
(06.05.2021/Munich) The Quantum Applications & Research Laboratory (QAR-Lab) at the Institute of Computer Science of Ludwig-Maximilians-University (LMU) continues to advance the field of quantum computing. It increasingly brings research knowledge into applications. On May 6 2021, Computer Science Professor Dr. Claudia Linnhoff-Popien, as head of the QAR-Lab, inaugurated the “Quantum Computing Optimization Challenge” as a two-month project of LMU and business partners that will compute industrial use cases on real quantum computers.
Founded in 2016, the QAR-Lab pursues the ambitious goal of making quantum computing (QC) accessible to a wide range of users in research and industry. Five major industry partners presented their use cases to have various optimization scenarios computed on quantum computers over the next ten weeks. A total of 288 interested parties had participated in the virtual kick-off.
The guest speaker at the kick-off event for the “Challenge” was Dr. Markus Hoffmann from Google Quantum AI, a division of Google Research. In the U.S., Google is building its own hardware that can be accessed through its cloud service. Hoffmann explained where quantum computing can be faster than a classical computer for an abstract problem. In doing so, he illustrated the October 2019 breakthrough when the U.S. company created a computer that calculated a sampling problem in just 200 seconds that would have taken a supercomputer 10,000 years.
Experts from BASF, BMW, SAP, Siemens and TRUMPF then presented their use cases. 27 students from the Institute of Computer Science will program the use cases on four quantum computers from May to June 2021 to find out what the quantum computers can already calculate, how complex tasks can be and how many QBits a company needs for its use case.
The speakers of the companies and their use cases:
- Dr. Astrid Niederle, Research, BASF about “Job shop scheduling” with test tubes in the laboratory.
- Lukas Müller, Group IT, BMW about “Pre-Production Vehicle Configuration”.
- Thomas Engelmann, Digital Supply Chain, SAP about “Beverage Delivery.
- Dr. Christoph Niedermeier, R&D Technology, Siemens about “Quedge“
- Frederick Struckmeier, Predevelopment Production Platforms, TRUMPF on “Production Planning in Sheet Metal Manufacturing”.
Claudia Linnhoff-Popien said: “The 20 quantum computing programs to be programmed will be run on four machines worldwide during the ten-week challenge and the results will be compared. We are very excited about our partner companies and about the joint project: such an extensive evaluation of real applications on four quantum computers is unique in Germany, maybe even worldwide.”
Use cases of the companies are focusing on optimization
The task at BASF in the area of laboratory research is to calculate how classic experiments in the laboratory can be carried out faster by changing the processes. The goal is to combine in which sequences robots have to bring which test tubes to which stations in order to achieve the fastest result. This is a simple case that becomes too complex for a classic computer, if – for example – it had to calculate 100,000 possible combinations.
In the task of BWM for„Vehicle configuration“ there will be optimized combinations of test components. In the process, components that are combined with each other in test vehicles should satisfy certain clauses so that as few vehicles as possible are required to test a given quantity of parts. After all, with an installed cable length of 10,000 meters, 100 million lines of source code and 10 60 possible combinations for one car, it becomes clear how complex special configurations can be for car orders.
SAP presented a use case with the “Bay Truck” in Beverage Delivery, which is intended to calculate the optimal supply deliveries of beverages in a special delivery area, when – for example – parameters such as delivery routes change. Here, too, it became clear how complex a daily delivery can become for a beverage company if the optimization affects 6,000 trucks per day.
Siemens presented a use case in the area of „Scheduling“. The aim is to calculate how certain tasks have to be processed one after the other in order to meet all deadlines. The variables here: short-term task changes, limited resources, new processes and new deadlines of the subtasks. Due to the short-term changes of several parameters, such scheduling calculations cannot be performed sufficiently fast on classical computers.
Trumpf’s use case looks at scheduling problems in sheet metal bending, welding, and painting. The goal is to optimize results when delays occur in production processes, for example.
Four solutions to one problem: Challenge finds best result in each case
In the Challenge, each problem is calculated and programmed on four computers (with two different computer architectures, the so called Gate and the Annealing model): this gives each problem four solutions. At the end, the performance of the computers and the quality of the solutions are compared to obtain an optimal result.
Prof. Dr. Linnhoff-Popien explains: “We want to find out which architecture calculates which result. To do this, we first have to make specifications. For example, in the production of sheet metal parts, the goal is to produce parts as quickly as possible or in parallel and to optimize the process. What is exciting for us is which architecture leads to which result and how stably, how scalable the tasks can already be executed on quantum computers today – and what requirement of QBits is necessary for the respective use case in order to achieve a quantum advantage.”
The Challenge serves to promote the transfer from science into practice: when completed after the Challenge, the results will be presented internally to the industry partners in July, before the results will be made publicly available as scientific publications.
QAR-Lab at the IT Institute has been working practice-oriented for years
The motto of the QAR-Lab is “Become Quantum ready”. For years, it has brought companies’ first use cases to the computers of the future. Claudia Linnhoff-Popien explains, “In our QAR-Lab – founded in 2016 – we have built up an enormous amount of know-how over the years to apply the technology of quantum computing in practice. Numerous well-known corporations are already benefiting from our knowledge.”
So far, the QAR-Lab is a unique place for students of LMU for practice-oriented events, in which – via the cloud – computing can be done on four quantum computers worldwide. Since 2018, university teaching has been geared towards testing quantum computing in a practice-oriented way beyond pure theory.
As a founding member of the outstanding European project PlanQK (“Platform and Ecosystem for Quantum-Assisted AI”), the QAR-Lab is also doing pioneer work by using quantum computing technology in the field of artificial intelligence. The experts of the QAR-Lab collaborate in the context of research collaborations on the implementation of quantum-assisted AI algorithms for industrial use cases.
Optimize and get faster: More companies launch pilot projects with quantum computing
Quantum computers, based on quantum technology (so-called Q-bits), can solve complex computing operations exponentially faster than previous computers and thus achieve a so-called quantum advantage, which will also translate into extreme speed of complex calculations. Estimates are that the hardware will be ready for the market in around five to eight years. Innovation-driven companies have long recognized the benefits of quantum computing. As a result, they are launching their first pilot projects in their IT or research departments in order to master the application of the new technology on the IT side in good time and make the technology commercially viable.
Speed is important to everyone: Advantages are expected, for example, in optimizing workflows, calculating complex processes or increasing efficiency and speed. In the future, it should be possible to calculate problems or scenarios within hours instead of months, within minutes instead of days. There are virtually no limits to the fields of application for quantum computing – whether in the pharmaceutical industry, the financial sector, logistics or the automotive industry. In the field of logistics and optimization in particular, there are virtually no limits, regardless of, for example, the optimal location of objects, the optimal sequence of processes, the optimal allocation of resources or the best combination of active ingredients.
Two models of quantum computers: Gate Model and Quantum Annealing
The range of possibilities is wide; quantum computers can be used to perform a wide variety of computational processes. Since the development of the hardware is not yet matured, it is not possible at this present moment to make a conclusive assessment of the extent to which one model is better than the other. The use cases of the „QC Optimization Challenge“ will be processed on four NISQ computers from the hardware manufacturers D-Wave Systems, Fujitsu, IBM and Rigetti and on two different computer architectures, the so-called gate and annealing models.
The quantum gate model is the quantum equivalent of the classical computer and in general is applicable to various problems. One of the most promising applications for the quantum gate model is material simulation. Current quantum gate models comprise around 50 qubits.
Quantum annealers, on the other hand, are specifically tailored for solving optimization problems.
The number of Qubits, like e.g. D-Wave Systems quantum annealer, are 10 times larger than that of the gate models (approximately 5000 Qubits).
However, the architectures are comparable only to a very limited extent, due to a different alignment and different fields of application. An evaluation – in terms of the applicability of different use cases to these different architectures – is being developed in the „QC Optimization Challenge“.
