Fakultät IuI
Refine
Document Type
- Conference Proceeding (6) (remove)
Language
- English (6)
Has Fulltext
- no (6)
Is part of the Bibliography
- yes (6)
Keywords
Institute
- Fakultät IuI (6)
Artificial intelligence (AI) and human-machine interaction (HMI) are two keywords that usually do not fit embedded applications. Within the steps needed before applying AI to solve a specific task, HMI is usually missing during the AI architecture design and the training of an AI model. The human-in-the-loop concept is prevalent in all other steps of developing AI, from data analysis via data selection and cleaning to performance evaluation. During AI architecture design, HMI can immediately highlight unproductive layers of the architecture so that lightweight network architecture for embedded applications can be created easily. We show that by using this HMI, users can instantly distinguish which AI architecture should be trained and evaluated first since a high accuracy on the task could be expected. This approach reduces the resources needed for AI development by avoiding training and evaluating AI architectures with unproductive layers and leads to lightweight AI architectures. These resulting lightweight AI architectures will enable HMI while running the AI on an edge device. By enabling HMI during an AI uses inference, we will introduce the AI-in-the-loop concept that combines AI's and humans' strengths. In our AI-in-the-loop approach, the AI remains the working horse and primarily solves the task. If the AI is unsure whether its inference solves the task correctly, it asks the user to use an appropriate HMI. Consequently, AI will become available in many applications soon since HMI will make AI more reliable and explainable.
With the increasing size and complexity of embedded systems, the impact of software on energy consumption is becoming more important. Previous research focused mainly on energy optimization at the hardware level. However, little research has been carried out regarding energy optimization at the software design level. This paper focuses on the software design level and addresses the gap between software and hardware design for embedded systems. This is achieved by proposing a framework for software design patterns, which takes aspects of power consumption and time behavior of the hardware level into account. We evaluate the expressiveness of the framework by applying it to well-known and novel design patterns. Furthermore, we introduce a dimensionless numerical efficiency factor to make possible energy savings quantifiable.
Knowledge of the maximum friction coefficient µmax between tire and road is necessary for implementing autonomous driving. As this coefficient cannot be measured via existing serial vehicle sensors, µmax estimation is a challenging field in modern automotive research. In particular, model-based approaches are applied, which are limited in the estimation accuracy by the physical vehicle model. Therefore, this paper presents a data-based µmax estimation using serial vehicle sensors. For this purpose, recurrent artificial neural networks are trained, validated, and tested based on driving maneuvers carried out with a test vehicle showing improved results compared to the model-based algorithm from previous works.