Proaktiivtehnoloogiate uurimislabori põhitegevuseks on võrgustatud küber-füüsikaliste-süsteemide õpetamine üliõpilastele ning teoreetiliste ja praktiliste probleemide uurimine/lahendamine – näiteks arendatakse küberfüüsikaliste süsteemide magistrikursust, lepingute toetusel ehitatakse ja rakendatakse spontaanseid sensorvõrke, teoreetiliselt uuritakse uduarvutust ja selleks kasutatavat proaktiivset vahevara (ProWare) ja osaletakse tervikliku olukorrateadlikkuse võimekuse uurimisel riigikaitse rakendustele. Lisaks teostab ProLab konvolutsiooniliste närvivõrkude abil uuringuid klassifitseerimise, semantilise segmenteerimise ja objektide tuvastamise kohta. Metoodikat on rakendatud fotopiltidele, punktpilvekogudele ja helisalvestistele. et
The laboratory focuses on the theoretical andpractical study of networked systems built fromstationary and/or mobile software-intensive(proactive) components. Typical componentsare pervasive computing systems. The researchis partitioned into three threads: (1) modelingand verification of situation-aware interactioncentered computation; (2) methods and technologies for acquiring situational information;(3) methods for processing and interpretation ofsituational information for (proactive) decisionmaking. The long-term goal of the laboratoryis the ability to detect and partially control the emergent behaviour in pervasive computing systems. In addition, ProLab performs researchon classification, semantic segmentation andobject detection using convolutional neuralnetworks. The methodology has been applied tophotographic images, point cloud collections andsound recordings.One of ProLab’s research focuses has been todevelop smart Internet solutions using energyefficient microcontrollers. Taking advantageof research in the fields of self-organization,multi-agent systems, cyber-physical systems,and interactive computing, we have shown thatit is possible to create very large sensor networksconsisting of distributed devices. The dynamicorganization of the cooperation between the elements of such a sensor network is very importanthere. One good example of this is the smart citysensor network on the streets of Tallinn, whichconsists of nearly 900 sensors designed and deployed by ProLab. ProLab is currently developingmachine learning models for microcontrollersfor the same sensor network that can detect different types of noise in urban environments aswell as the types, direction and speed of vehiclesmoving on the streets. en