Horizon scanning 2019
Publish date: 2020-10-28
Report number: FOI-R--4995--SE
Pages: 16
Written in: Swedish
Keywords:
- Technology foresight
- horizon scanning
- literature study
- research front
Abstract
This report provides a summary of horizon scanning activities performed by FOI in 2019. The aim is to identify research advances that may be particularly relevant for the Swedish Armed Forces, and to increase FOI's overall knowledge of new technologies. For each topic, a literature survey was conducted to evaluate its status and relevance. A brief description of the five studies carried out in 2019 is given below. A quantum radar can potentially increase low-probability of intercept radar functions, decrease the sensibility to electronic warfare, increase possibility to detect stealth vehicles and increase target recognition capabilities. A quantum radar would be robust to noise and effect losses. We estimate that it would be possible to develop a quantum radar in the near future. The first applications will likely be low-probability-of-intercept radar followed by detection of stealth vehicles. We studied the research field on artificial intelligent cyberattacks and how the technology could be applied in the different steps of an attack. The study focused on research from an antagonistic perspective and on technologies that can be used for both attack and defense. Nineteen use cases of artificial intelligent cyberattacks were identified within the different steps of an attack. Some of the use cases, and the underlying AI technology that is employed, indicate that some parts of the field have reached maturity. The on-going development of deep learning has opened new possibilities for synthetic manipulation in video and voice recordings, so-called deep fake. Literature and recent events were studied, especially technologies for creating deep fake, automatic detection of deep fake, and possible countermeasures. The study discuss how deep fake, with a small effort, can create disinformation and make evidence in the form of video and voice recordings questionable. The area is developing rapidly and scientific results are converted into software that can be applied by a user without deeper technical knowledge. There is an inherent limitation in the currently developed machine learning methods. These methods focus on modelling of correlations in data. By including also learning of causal structures, it will be possible to enhance understanding and reasoning of events. Today there are examples of civilian applications in risk analysis. We conclude that learning and detection of causal structures will be of importance in future military system that include machine learning. The human body contains at least 100 trillion microorganisms, so-called microbiota, where the vast majority are found in the gastrointestinal tract. The genetic material of the human microbiota is known as the human microbiome. Recent years' research has shown that the microbiome is involved in several essential functions in the human body. Malfunctions in the microbiome can lead to impaired health, for example inflammations, increased sensitivity for infections and diseases, decreased cognitive functions or development of chronical diseases. The rapid technology development within gene sequencing and AI will further increase the speed of this research field. However, more research is required before the true connection between healthiness and illness and the human microbiome can be clarified.