Research and Education Goals:

Rapid advances in sensing, computing, communication, machine learning, artificial intelligence, statistics, brain sciences, and allied fields are paving the way for the realization of complex embedded, distributed, intelligent, and networked information technology (IT) systems and platforms. Our long-term research focus is to advance the development of a ‘design theory for artificial intelligence systems’ and to build automation tools and engineering platforms that will allow for systematic design of intelligent systems. Addressing this crucial goal will enable the efficient and systematic design of complex IT and intelligent systems of the future that are needed to address world’s mega-challenges that society faces – e.g. ageing demography, climate change, migration to cities and overpopulation, etc. Our teaching goal is to help prepare future scientists and engineers who can: a) work in large teams and projects, b) lead and manage distributed, global (multi-cultural) teams, c) have depth in critical technical areas as well as sufficient breadth to solve systems problems of interdisciplinary nature that is needed for the future, and d) are “Holistic Systems Thinkers”. Our current research focus is on model-driven design and engineering of complex intelligent systems and in interdisciplinary research bridging systems engineering, computational neuroscience, cognitive science, psychology, AI, and statistical machine learning.

Philosophy – Systems Engineering for Intelligence:

The essence of our viewpoint on computer vision system design is that: visual cognition involves context-dependent, hypotheses generation followed by detailed estimation via deliberation or iteration. This view is analogous to the dual-system model articulated in the psychology literature (i.e. The brain processes can be divided into two systems: system 1 – subconscious processes of the brain that use fast heuristics to generate hypotheses, and system 2 – processes that deliberate on system 1’s output, and uses declarative facts and rules to come to conclusions). The systems engineering design framework thus maps formal models of application contexts, sensor types and configurations, task descriptions and performance specifications to dual-system architectures and respective modules. Our current vision architecture work is still far away from a biologically plausible implementation and can be viewed as technical systems developed via influence from biology. A key long term research issue is to identify how a brain-like architecture with homogeneous components along with specific learning mechanisms can lead to a range of intelligent functions and behaviors. The feedback from lessons learned from artificial system designs to brain sciences and back will enable better understanding of how to build general artificial intelligence systems and may give insights for hypotheses on functional theory of the brain.


Concrete results from the Bernstein Focus NeuroTechnology – Frankfurt, EU H2020 project AEROBI  initiatives  include the fusion of model based systems engineering with advances in neuroscience and machine learning to develop platforms and applications involving cognitive vision. The results of these efforts are actively translated to industrial applications in logistics and mobility through a research cooperation with Frankfurt House of Logistics and Mobility GmbH (HOLM) and their industrial research partners.  Ongoing projects include BMBF funded ‘AISEL: AI Systems Engineering Laboratory’, EU H2020 project RESIST.

Additional details can be found in the following posters:

  1. Systems Engineering for Intelligence  (Posterv4)
  2. Engineering Platforms for Cognitive Vision and a Case Study (poster-sys)
  3. Simulation for Cognitive Vision  ( wacv_poster )


Prof. Ramesh Google Scholar Page

Prof. Ramesh DBLP Page