Automatic recognition of manual activities – thanks to the intelligent combination of object and gesture recognition with increased precision and ready for industrial use!

MAMOC in action

To visualize the progress with MAMOC in more detail and to demonstrate the showcase, our developers have created the following video (German):


Mit dem Laden des Videos akzeptieren Sie die Datenschutzerklärung von YouTube.
Mehr erfahren

Video laden


MAMOC is a funded EU research and experiment project in the context of the MIDIH (Manufacturing Industry Digital Innovation Hubs, LINK). The aim of MIDIH is to provide digital innovations and services for the manufacturing industry with particular focus on the needs of small and medium sized enterprises (SMEs).

We are proud to be part of this and to make our own contribution.


MAMOC (Machine Learning Application for Motion Capture)

The aim is to enable precise and detailed action recognition for the industry. The application can be particularly useful at manual workplaces and can support topics related to time management and industrial engineering.

Even highly complex production processes in the manufacturing industry are subdivided into individual work steps organized according to the process. In its most rudimentary form, every single work step has a clearly defined beginning and end and a specified working time.

The total of the working times for all individual work steps of a product determines – to a considerable extent – the total process time required to manufacture a product. As a consequence these process times are also a significant factor in calculating production costs.

When it comes to fully automated products, process times can be recorded precisely to the second, but they are considerably more difficult to determine for manual activities. This is where the automatic action recognition of MAMOC comes in.

With MAMOC we create an objective data collection with sufficient accuracy for several use cases, answering the following questions:

  • How long does a single work step take?
  • How long do my work steps take on average?
  • Has a work step been performed correctly?
  • Have all work steps been carried out in the correct order?

As soon as work steps are defined and breakdown into activities is done, MAMOC can help to answer these questions in a statistically significant way. Each of the automatically recognized activities is labelled with a precise time stamp and thus helps to provide the statistical basis for calculating average values for the individual work steps.

What lies behind MAMOC?

Vision & AI
Gesture Caption

MAMOC is based on the intelligent combination of the AI technologies object recognition and hand gesture recognition.

Following an initial training, the system is able to recognize the trained manual activities and automatically records their working times.

The current showcase includes: hammering, screwing and cutting.
Certainly, sufficient training will more and more enable recognition of other activities and entire work steps.

Everything required on the hardware side can be provided as embedded solutions (e.g. from the manufacturer Nvidia).
Unlike other AI applications, it does not require a strong gaming PC or cloud-based processor power. MAMOC is exclusively operated with specializedlightweight embedded hardware without an active Internet connection.

Together on to new horizons

MAMOC is exactly what you were looking for?

Maybe you already have the next challenge for our MAMOC in mind?

Either way – we help you to find out if and how you can use this AI application for your benefit!

Our team is ready for you – mail to or call (+49 89 427 74 177).