You could say it's in our DNA since birth.
Today several of our employees have a PhD and many years of experience from working in research organisations and academia which enables us to now drive our own research projects as well as being a research partner.
With our combined experience from research within the fields of machine learning, distributed computer systems, organisational theory and distributed product development it is essential for Substorm to continue to be a part of, and drive, research projects together with both customers, partners and universities across the globe.
If you are interested in taking part in a research project or find out more about one of the ongoing projects don’t hesitate to contact us.
Ongoing Research projects;
- NoBias (LTU, Teknikkvinnor, Tromb)
- Future Trust (LTU)
- Circular product design for small and medium sized producing enterprises (Sepro AB, Sannineri AB, Wanjet AB)
- NLP for Customer Service/Service Delivery
- Predicting company culture using
My research started in e-health where we studied how to slow or prevent the onset of dementia using a wearable device. Sensor data from the device together with machine learning was used to make activity recognition. At this time (and still is to some extent) computational power was a limiting factor in order to run machine learning on the device itself, so data needed to be uploaded to a more powerful computer. Further, many studies used different sensors placed on different parts of the body. My work focused on how to perform machine learning on resource-constrained devices and also considered how to make the sensing as unobtrusive as possible.
This was in order to take this technology from a laboratory setting towards an end-user application
- Unobtrusive Activity Recognition in Resource-Constrained Environments
- Activity recognition in resource-constrained pervasive systems
- Time-efficient algorithms for laser guided autonomous driving
- Classifier Optimized for Resource-constrained Pervasive Systems and Energy-efficiency
- Low-Power Classification using FPGA: An Approach based on Cellular Automata, Neural Networks, and Hyperdimensional Computing
- A Domain Knowledge-Based Solution for Human Activity Recognition: The UJA Dataset Analysis
I started of my research career in the field of integrated product development in manufacturing industry. The research focused on key parameters that enables increased efficiency and early integration of organizational and technical capabilities, for instance the use of applied machine learning and how trust factors could be a mitigating factor. 15 years has past since then but the area is still very topical and the challenges in the industry still exist. That is also why we have an extra focus on the area in our research project FutureTrust. Equality and inclusion is another area that hos been very close to my heart for many years and that goes hand in hand with performance in the industry. There we work together with LTU and Teknikkvinnor in the project NoBias. NoBias aims to prevent exclusion with the use of machine learning. Last but not least the area of transforming traditional business models from product sales towards cloud based services and revenue share models and how automation enables this, is an area of great importance today and in the future. Here I would like to see more applicability and thus actual test and experiment on such models to be able to understand the opportunities that lie ahead.
- Important Factors for Project Performance in Collaborative Product Development: A Survey Investigating Contextual Settings
- Managing Collaborative Product Development: A Model for Identifying Key Factors in Product Development Projects
- Enabling Knowledge Transfer in Product Development and Production: Methods and Techniques From Artificial Intelligence
- At a Crossroads: Case Study Analysis of the Organizational Challenges within the Transformation Path to an IPS2
- Whats in it for the Provider? The Case of a Telecom Vendors Value Capturing from the Transition to Product-Service Systems