CooTSIoT Project


HW-SW co-design and optimization of Time Series based applications for IoT ultra-low power embedded devices (CooTSIoT)

Project funded by the Spanish State Plan for Scientific, Technical and Innovation Research for the period 2021-2023, within the framework of the Plan for Recovery, Transformation and Resilience (PRTR).

Project data

  • Title: HW-SW co-design and optimization of Time Series based applications for IoT ultra-low power embedded devices
  • Acronym: CooTSIoT
  • Project ID: TED2021-131527B-I00
  • Funded by: MCIN/AEI /10.13039/501100011033 and by European Union Next GenerationEU/ PRTR
  • IP and Co-IP: Angeles Navarro and Rafael Asenjo
  • Duration: 1/12/2022 to 30/11/2024: 24 months
  • Budget: 233.680,00€

Summary

CooTSIoT will develop new personalized healthcare and smart agriculture monitoring applications and services by leveraging computational resources located at the Edge under the IoT ecosystem. We will take advantage of synergies with project collaborators that are either experts in biomedical applications on wearables or specialists in the study of the impact of water and pathogen stress in subtropical cultivations, along with experts in the development of software and hardware technologies to develop novel ultra-low power (ULP) battery-based embedded systems able to process sensor data on the same device. Therefore, these systems will effectively bring the extraction of knowledge from the Cloud to the Edge, closer to the sensors. To move the services and processing of data to the Edge we will use disruptive technology covering almost all levels of description of a system: in the algorithm level we will innovate with state-of- the-art big data analytics approaches such as Contrast Profile to design a novel anomaly detection method for IoT devices that process time series from sensor data; in the system software level we will pioneer the use of lightweight programming models based on the parallel task paradigm for our targeted ULP power systems; we will design and exploit advanced runtime techniques to best leverage the hardware resources that embeds the applications while minimizing energy consumption; and in the system hardware level we will design and optimize a next generation ULP power System-on-Chip (SoC) device based on the emerging RISC-V architecture, producing devices that run our applications below ~10 mW.

The personalized healthcare monitoring application embedded on wearables compatible with a normal social life, will help to improve the quality of life and control the occurrence seizures in patients affected by epilepsy, that is one of the most common neurological disorders affecting around 65 million people worldwide. The smart agriculture monitoring application will help to identify in real time, anomalies that affect crop/plant irrigation and their effect during fruit development and ripening in subtropical plantations (strawberry, avocado, mango), as well as variations in the plants due to the use of pesticides and chemical treatment. Therefore, our project results will contribute to some relevant environmental objectives of the European ecological transition initiative: the smart agriculture monitoring application can be designed to minimize the use of water resources of the crop/plant irrigation system in order to ensure optimal production; it will also help farmers to better adjust to climate change and ensure the sustainability of the cultivation processes; but also, the smart agriculture monitoring application can be tuned to measure the effect that chemical treatment or pesticides have on the behavior of pests, allowing the farmers and experts to reduce in real time their use to control water and soil pollution, effectively minimizing its ecological impact.