Industrial Ph.D. students
Within WASP at KTH, there are five industrial Ph.D. students. Besides the work they do at KTH, they hold positions at ABB and Ericsson.
Project: Disaggregated Data Center (Ericsson)
Amir Roozbeh (Ericsson) will be looking at new data center (DC) and how infrastructure architectures break the physical server-oriented model via hardware disaggregation.
Disaggregation means that each type of resource in the DC is seen as a resource pool. Logical servers are composed of a subset of
the DC’s resources interconnected via the DC's fast interconnect network. This approach brings greater modularity, flexibility, and
extensibility to DC infrastructures, allowing operators to employ resources more efficiently.
Project: Software Architectures for Autonomous Energy Efficient Resource Allocation in 5G RAN (Ericsson)
Diarmuid Corcoran (Ericsson) will investigate the impact of new, distributed, software techniques on the energy efficiency of 5G radio access compute infrastructure. The golas is to create intelligent software techniques that will contribute to affordable, sustainable and energy efficient 5G radio access networks in a virtualized and cloud context. A fundamental principal will be the mechanisms and techniques to enable autonomous energy efficient architectures and algorithms
Project: Optimal Scheduling in Underground Mining (ABB)
Max Åstrand (ABB) will be focusing on underground mining. Underground mining comprises a range of activities, from rock excavation tasks (such as drilling, charging and blasting) to support functions (including managing the inflow of water, ventilating blast fumes, and building the necessary infrastructure). Automation in underground mining has previously focused on fixed equipment such as mine hoists, crushers, and conveyor belts. For underground mines to reach high productivity automatic scheduling of the mobile production system must be considered.
Project: Skill Acquisition for Industrial Robots (ABB)
Shahbaz A. Khader (ABB) will look at how industrial robots can acquire “skills” which can then be employed in a production scenario such as an automobile assembly line. A human worker develops such a skill after receiving formal instructions followed by some practice. Nobody tells the human how to move his hands exactly. Robots, on the other hand, have to be meticulously programmed to produce the exact motion which would result in the job getting done. This programming phase is expensive and time consuming. So we ask the question: Can a robot, provided with high-level instructions and/or shown a demonstration, practice on its own and acquire an optimal skill?
Project: A Supervised Learning Approach to Fast Link Adaption (Erisson)
Vidit Saxena (Ericsson) project will focus on:
LTE networks adapt the transmission parameters for each link according to the instantaneous state of its radio channel. One aspect of this so-called Fast Link Adaption (Fast LA) is to predict the optimal modulation and coding scheme (MCS) on a per-subframe basis, subject to a desired long-term Packet Error Rate (PER). In this report, we design and study neural networks for predicting the MCS for Fast LA in the LTE uplink. Our simulation results show that the neural network is able to learn the uplink receiver characteristics for a simple single-link scenario, and predict the optimal MCS under the specified PER constraints.
Project: HDV Precise Maneuvering in Urban Environments
Rui Oliveira (Scania) will look on the fast development of Autonomous Driving that is soon bringing automated vehicles into city roads. Urban driving is characterized by fast paced, tightly spaced and complex maneuvers, which add layers of difficulties autonomous cars must deal with. Moreover, when considering Heavy Duty Vehicles (HDVs) such as trucks and articulated busses, extra complications arise due to their large dimensions, slow dynamics and multi-body characteristics. New solutions must be sought which build upon existing techniques and extend them to vehicles of this nature.