Our AI Showcases

ARIC Showroom in Hamburg

By using artificial intelligence (AI), companies can achieve higher efficiencies, make more informed decisions and develop completely new business models. With pre-trained solutions as a cloud service, i.e. an "AI at the push of a button", so to speak, business-generating AI can already be used today, especially for smaller companies.

Especially in the field of image processing, massive progress has been made in recent years, which has led to this AI technology being used in very different areas and has already greatly changed the world of work.

Its work focuses on the initiation of pilot projects, the interdisciplinary application and dissemination of AI-related know-how, and the promotion of research and development cooperation between business, science and society. The main work areas of the competence center are the initiation of a comprehensive knowledge transfer in the field of AI between business, science and society as well as the provision of interdisciplinary target group specific event and educational formats, and the initiation and support of application-oriented R&D projects. A nationwide exchange with AI experts, centers and multipliers takes place on a regular basis in order to track down current trends, developments and best practice examples and to promote and network the AI ecosystem in Germany.

Our AI showcases in detail

Baggage Control AI - Baggage Conveyor System.

The demonstrator shows the local use of an AI on a miniature high-performance computer, which is able to evaluate live the loading of baggage trays (suitcases, backpacks, etc.) on a baggage conveyor system with regard to problematic overhangs.

But also general product defects, e.g. faulty solder joints or wrong arrangements or bearings, can be automatically detected by the AI already in real time. The complexity here is particularly high, since a wide variety of unknown objects must be continuously analyzed in real time. Overhanging baggage tags, for example, do not cause any disruption.

The AI is trained for this in advance or also regularly in the cloud. No Internet connection is required to operate the AI.

Insights on the showcase

The baggage handling system is the nervous system of the airport; suitcases must be reliably loaded from the baggage counter to the plane and vice versa
Baggage conveyor systems of large airports or cruise terminals have several kilometers of conveyor line and are difficult to access in many places
After the suitcase disappears behind the check-in counter, it lands in a tray (luggage trough) without being fully automated and begins its journey
The vast majority of baggage thus finds its way to the loading area without any problems
Rarely do banners or straps hang out of the tray or whole pieces of luggage protrude over the edge of the tray and can cause sensitive disturbances
Such protrusions and overhangs can get caught in the system and cause a piece of baggage to be ejected from the tray, block the system, damage it or activate the emergency stop
Baggage tags are an exception, they often overhang due to their attachment but rarely cause malfunctions.
As a preventive measure, all trays should be evaluated shortly after loading for possible overhangs and overhangs, so that these are detected at an early stage and routed to a point where an employee can ensure proper tray loading. Late tags are to be intelligently detected and ignored in the process.
This showcase demonstrates how such an evaluation can be implemented in a reliable, traceable and economical way.
A camera mounted above the conveyor belt delivers images in a high frame rate to a system with embedded AI, a small high-performance computer, which evaluates them in a two-stage process - In the first stage, it is detected highly efficiently whether a tray is centered and reasonably central under the camera
If a tray has been detected, the tray is evaluated in the second stage. This works via segmentation, each area of the image is assigned by the AI either to the tray, the piece of luggage, the luggage tag or the background. The result is a mask on which it can be seen whether a critical part of the luggage survives. In this case a warning message goes out.
No matter where objects move on conveyor belts or other transport routes (e.g. on rails or fairways in the port), they can be detected and analyzed. In this way, correct loading positions, loading quantities or goods classes can be determined, tailored to specific use cases.

Visitor Prediction AI - Visitor Flow Analysis

The demonstrator shows a Visitor Prediction AI that counts people on video streams and helps to direct visitor flows intelligently.

Image processing can be used not only for objects or product analysis, but also for anonymized analysis of people - for example in danger zones. Where are people located and which of these people are motionless and need help? This would be possible, for example, for the rapid rescue of drowning victims using appropriate image recognition processes. Here, too, compliance with the GDPR and the protection of privacy are essential.

AI applications thus even make it possible to save human lives. Image processing solutions for intelligent visitor flow analysis have also proven themselves in the Corona era, for example in our projects in St. Peter-Ording and Scharbeutz, where our solution was awarded the German Tourism Prize 2020. You can find more information here.

Insights on the showcase

Possibility of forecasting visitor flows
Users learn how many people are currently at a location (in the showroom, on a beach, or in the airport) and how this number is likely to develop in the near future.
Advantages e.g. at tourist hotspots: satisfied tourists and thus recurring revenues; marketing of advertising space at visitor hotspots
Artificial intelligence (AI) as a service from the cloud: use AI now as AIaaS (Artificial Intelligence as a Service) from the cloud
AIaaS: AI as a service refers to turnkey, out of the
How does this work? Through application programming interfaces (APIs), users can harness the power of AI without having to write a single line of code
Advantages: Flexibility (customization of the service to meet individual business needs), Transparency (customers only pay for what you use), Scalability of AIaaS depending on a company's needs and growth, Cost-effective (outsourcing of AI solutions), Quick and easy to set up (without a team of experts and complex infrastructure on the part of the company)
A camera mounted in the room delivers the video stream to a small embedded computing unit
A button linked to the small embedded computing unit takes a photo of the room and sends the photo to Visitor Prediction AI in the cloud - The service (Visitor Prediction AI) runs the AI in the cloud and the result is displayed on a dashboard
The dashboard shows the live stream of the room, the photo with the detected people (surrounded by a bounding box that also anonymizes the people - AI without sacrificing stability and security), the analysis of the photo with the number of people, and a graph of the history of the number of people counted in the showroom.
Visitor Prediction AI shows how our AI services enable rapid implementation without sacrificing stability and security, which includes many other deployments: Visitor measurement, intelligent visitor flow management or Corona security compliance.

Downloads

Brochure Artificial Intelligence
pdf, 873 KB

What is Artificial Intelligence? What should a good AI be able to do? Where are their limits? How does an AI learn? And various other aspects are covered in our brochure on Artificial Intelligence. Download the brochure here.

Whitepaper „Artificial Intelligence as a Service“
pdf, 2 MB

Read all about the benefits of AIaaS HERE and learn how AI projects can be implemented quickly and easily, even if you have no infrastructure of your own.