Do green computing better
What is Climate-KIC?
EIT Climate-KIC is a European knowledge and innovation community, working to accelerate the transition to a zero-carbon economy. Supported by the European Institute of Innovation and Technology, they identify and support innovation that helps society mitigate and adapt to climate change. Climate-KIC’s vision is to contribute to creating a prosperous, inclusive, climate-resilient society with a circular, zero-carbon economy. We believe that innovation will achieve the deep decarbonisation required, and strengthen climate resilience.
Climate-KIC brings together partners in the worlds of business, academia, and the public and non-profit sectors to create networks of expertise, through which innovative products, services and systems can be developed, brought to market and scaled-up for impact. The Climate-KIC Accelerator is the only EU acceleration programme focused on climate impact by cleantech commercialisation. Organized in three stages, the accelerator is an 18-month program bringing the knowledge, resources, tools and the coaching a cleantech start-up needs for success.
How did Climate-KIC support our venture?
As a young and enthusiastic start-up, we first applied for the Accelerator program, stage 2, in 2017, then again in 2018 for stage 3, and were accepted with our project “Low-Carbon Mobile Computing” which focused on using smartphones for efficiently communicating and collaborating towards reducing the energy footprint of entire communities.
We proposed a mobile-to-mobile contextual offloading model to promote energy efficiency,. This is a novel collaboration solution based on a contextual search algorithm which schedules the remote execution of tasks in community by predicting the availability and mobility of nearby devices (including powerful set-top-boxes at the edge of the cloud).The product we are developing implements this model as a platform on top of which applications can be developed. The heart of our solution is the contextual search algorithm based on the feature prediction model, which is responsible for deciding whether to offload a task on a nearby device or execute it locally. To be able to schedule tasks in opportunistic networks, the contextual search is split into workload profiling (ascertaining the resources needed to run a task) and the offloading decision (determining if a task is better run locally or on a different device). As any task can run on multiple devices with different traits, workload profiling needs to determine which resources are required by the task, compute a workload score which uniquely describes the computational requirements of the task, and normalize the workload score tobe able to compare the executional needs of a task for various platforms.
We are grateful to have had the honor of being selected in the Accelerator program and to have participated in the numerous trainings and coaching session offered by Climate-KIC. We consider this to have been origin story for one of our main products, Opportunity.