SIEMENS

After traveling the back country of Upper Austria in trains, trams, and buses, the trip ends in a landscape as noted for beauty as for brains. Here, in the town of Hagenberg, a mere 30-minutes’ drive northeast of Linz, an important segment of Austria’s software elite has set up shop. Otherwise known as “Silicon Hills,” this remote location, is a magnet for Dr. Georg Bodammer from Siemens Technology Accelerator GmbH (STA), who travels here from Munich many times each year. Bodammer’s job at STA – a wholly owned subsidiary of Siemens AG and part of the Corporate Technology department (CT) – is to evaluate new business ideas and help entrepreneurs to establish and develop their startups. And Hagenberg’s software park is a hotbed of ideas.

The park is home to 71 small companies and houses the University of Applied Sciences of Upper Austria and its approximately 1,500 students. Siemens also has close ties with Johannes Kepler University (JKU) in Linz, where 17 institutes work on IT and software solutions, often in collaboration with researchers from Siemens Corporate Technology.

One of the Hagenberg software park’s up-and-coming tenants is Arelion, a startup that specializes in optimization. Housed in the park’s most modern building since March 2012, the company is headed by co-founders Dr. Peter Feigl, Dr. Norbert Lebersorger, and Dr. Thomas Scheidl, who collectively own around 75 percent of the company; STA owns just under 25 percent.

The entrepreneurs developed their technology while working as scientists at JKU on a research project funded by CT – and they haven’t slowed down since. “Often, two of our people are on the phone, two are discussing a problem, and one is writing a program in the middle of all the noise,” says Lebersorger. If things go as expected, Arelion will hire ten additional employees this year – and will therefore have to find new offices. The company’s prospects are good – which is one reason why Bodammer is in Hagenberg so often. “I believe it’s very likely that Arelion’s business idea will succeed on the market,” he says.

“Our goal was to develop innovative optimization software,” explains Feigl, who, with his colleagues, spent ten years developing solutions that he says are faster and more flexible than established optimization methods, a view that’s been confirmed by CT.

Most traditional optimization techniques are based on linear programming methods, which means that a list of computing operations are performed until the optimal result has been found. Such a method is used, for example, if production operations need to be planned as quickly or as cheaply as possible, or with a certain number of machines. The result achieved cannot be mathematically disproved – in other words, there is no better solution. “Still, the calculations can take a very long time,” Feigl explains.

The young entrepreneurs’ alternative is to utilize a heuristic approach in order to achieve the best possible result – rather than a theoretical optimum – as quickly as possible. This enables program users to react very rapidly and flexibly to changes, particularly since Arelion’s software doesn’t require mathematical formulas for the modeling process. In addition, its object-focused programming makes it possible to flexibly depict even the most complex processes. The researchers demonstrated their software to Siemens using a program that controls a robot in a printed circuit board manufacturing process. The software calculates the movements the robot arm needs to make in order to mount electronic components on a board as quickly as possible. Arelion’s technology quickly finds a very good solution by eliminating inefficient paths of movement.

“When running on a powerful computer, the software can check results a hundred thousand times per second,” says Lebersorger. Because it operates so rapidly, the solution is particularly suitable for use with processes in which conditions change frequently. The transshipment center of an Austrian automotive logistics company offers a case in point. The center, which can accommodate 12,000 cars, delivers vehicles from various manufacturers to associated dealerships. Cars are constantly entering and leaving the facility. In the past, employees drove the cars onto an expansive site accompanied by a minibus that transported them back to the office. The same procedure was carried out the other way around when an employee had to pick up a car for delivery to a dealer.

“This was extremely inefficient,” says Feigl. With this in mind, Arelion developed a completely new system for the logistics center. This solution electronically registers each new vehicle that enters the facility. The system depicts every vehicle and its location, and every shipment date and time. Virtual experts optimize individual operations. All employees carry a mobile terminal that shows them their given assignment and guides them through the work day.

If, for example, a car is ordered on short notice, the program determines which employee is in the vicinity and sends him or her to the vehicle’s location. Minibuses are no longer needed because employees always drive onto the lot with one vehicle and leave with another. Thanks to better utilization of the workforce and more rapid completion of assignments, the company’s productivity has greatly improved.

Optimization in the Cloud. Arelion’s founders are convinced that the versatility of their technology is a key advantage. Applications include product planning; communication network optimization; smart grids; material flow, packaging, and logistics planning; and layout optimization for entire production facilities. Arelion is also working on an optimization-in-the-cloud service that will enable it to offer its solutions to a broad range of users. With this service, it will be possible to use computing capacity more effectively to manage more complex tasks. The service will also be offered with a pay-per-use invoicing system. As a result, smaller companies will be able to optimize their systems at a reasonable price. For instance, a carpentry workshop would be able to calculate the most effective and cost-efficient way to cut wood for a particular job, while a repair shop would be able to generate optimized repair schedules.

Arelion’s optimization specialists are looking even further ahead with systems such as an Internet platform for addressing all types of planning problems. “Users would enter the jobs they need to manage and optimization specialists would offer solutions that could be implemented with our platform technology,” Feigl explains. Such a service would be of interest to any company unable to afford major software investments, since an implemented solution could be reused by other companies facing a similar situation. In other words, any company that has joined the platform could access an appropriate existing optimization solution and run it with their own data. Who knows? Maybe one day Austrian transport companies may ask Arelion to help improve the connections between the Hagenberg software park and the world beyond the beautiful countryside.