Embedded Intelligence (Evaluated 1.4 in SoSe17)
After finishing this lecture the participants will have the basic understanding on artificial intelligence, realized in the format of compact wearable or ambient embedded systems, that automatically recognize and smartly response to the context of a single person or a crowd without interrupting their lives. This lecture focuses on the basic technologies of Embedded Intelligence, e.g., to acquire information from human beings and the environment, to build models of them and applications on top.
  • Questions, problems to be solved and examples
  • The attributes and application areas of different sensors
  • Different signal processing and machine learning methods for different activity recognition tasks
  • Key points to be considered in building the architecture for activity recognition
  • Dynamic sensor configuration
  • Performance evaluation
Lecture slides
  1. Introduction
  2. Location
  3. Activity
  4. Activity with objects
  5. User State
  6. Power
  7. Summary
  8. Summary-2
Sensors (Evaluated 1.4 in WiSe16/17)
After finishing this lecture the participants will have the basic understanding on sensor systems (Low level signal processing and digitization). This lecture gives an overview of sensors with a physics and electrical engineering perscpective. Computer Science and IST students can improve their understanding of hardware systems. Which can provide a better overall performance of the embedded systems.
  • Physical models of the sensors
  • Electronic models of the sensors
  • Digitization (ADC, DAC)
  • Analog signal processing
  • Low-level digital signal processing
  • Noise and filter
  • Different sensors and their applications
Lecture slides
  1. Sensor and its characteristics
  2. Physical principles of sensing-1
  3. Physical principles of sensing-2
  4. Interface electronic circuits-1
  5. Interface electronic circuits-2
  6. Sensor Technics-1
  7. Sensor Technics-2
  8. Sensor Technics-3
  9. The Lab for Sensors
Ubiquitous Computing Lab (Praktikum)
After completion of the lecture, the students will have opportunity create an entire application chain of one or more current Ubiquitous sensor systems figure out the necessary design factors. Basic methods and algorithms of activity detection will be applied in practise.
  • Embedded platform
  • Data Mining for activity detection
  • FPGA and / or microcontroller programming
  • Wired and / or wireless data transmission
Prof. Jingyuan Cheng
Teaching Assitant (WiSe17/18)
Yujiang He, Zhixing Huang
  1. Introduction
  2. Force sensor matrix system
  3. Python Framework for resistive textile sensing
Teaching Assitant (WiSe16/17)
Ankur Mawandia, Heber Cruz
  1. Force sensor matrix system
  2. Unified Matlab Framework for resistive textile sensing -1
  3. Unified Matlab Framework for resistive textile sensing -2
Presentation from participants
  1. Pressure password : robust personal identifcation using hand pressure
  2. Smart pillow cover : sleeping posture identification
  3. Spy on me : workspace behaviour identification with arm positions
  4. Smart stuffed toy :Identifying child interactions with a toy