Aftos Decision & Control
On-the-fly management, automated control and optimisation of data for machines
What is Aftos?
Aftos Decision & Control is a modular onboard AI system which utilises our patented APACC control architecture (see also Artificial Precognition).
Aftos harnesses the power of explainable AI (XAI) to learn how a system, for example a vehicle, behaves and provide robust (risk-averse) and adaptable control with a significant level of inferential thinking and real-time communication.
Traditional control methods make a lot of assumptions about the system and ignore uncertainties to simplify the design process; this results in degeneration of controller performance. Aftos takes uncertainties and non-linearities into account and produces robust and precise control, even in new (“unseen”) scenarios.
Aftos provides a controller which is easier to develop and deploy compared to traditional methods in avoiding the complicated process of designing a detailed model of a system. Furthermore, it can adapt to gradual changes in system dynamics (which for instance can be caused by actuator degradation), and therefore always provides the best possible control. Aftos can also make use of environmental information, such as the weather, which is often a source of uncertainty, by using it as one of the inputs - allowing for more comprehensive prediction of outcomes.
Case Study: Identifying Cancer Cells with Machine Learning Using Oscar
The early diagnosis and prognosis of a cancer type has become a necessity in cancer treatment. Massive Analytic Limited developed new algorithms for use in precision medicine for personalised treatment of cancer victims.