Industrial Doctorate HIMIKODE and CiTIUS on collaborative industrial robotics

CiTIUS is a research center specialized in Intelligent Technologies with a team of more than 120 researchers. Our team is the greatest asset of our center, and we are constantly looking for scientific talent. CiTIUS offers a stimulating interdisciplinary scientific environment at the cutting edge of research, where our researchers can enhance and develop their scientific careers within an international team. We are part of a high-level scientific environment with five research centers specialized in different areas (CiQUS, CiMUS, IGFAE, CRETUS, and CiTIUS). Santiago de Compostela is a UNESCO World Heritage City with more than 5 million square meters of green spaces, and it was named one of the 100 best places in the world by TIME magazine in 2021.

The University of Santiago de Compostela (USC)

The USC offers a vibrant and stimulating environment for research, characterized by academic excellence and a rich history. Founded in 1495, USC is one of the oldest universities in Europe and is recognized for its significant contribution to science and culture. With approximately 20,000 undergraduate students and more than 2,500 doctoral students, USC provides a dynamic and diverse academic community. The university has been awarded the 'HR Excellence in Research' by the EU under the HRS4R process, highlighting its commitment to supporting researchers.

Himikode

Himikode is a technology-based business initiative supported by the Xunta de Galicia, and it currently holds the 'Innovative SME' seal from the Ministry of Science and the 'Emerging Company' status granted by Enisa. Our activity focuses on creating innovative solutions that combine IoT, artificial intelligence, Big Data, and SaaS, developing products to create scalable businesses. From designing our own IoT electronics and developing AI models to delivering final software to clients, our mission is to drive digitalization and efficiency in key sectors like the food industry, food services, and energy. Key products include hkoffee and RuralData, which optimize processes and improve sustainability.

Position Description

Candidates will apply with CiTIUS and Himikode to the Doutormentos Industriais call from the Xunta de Galicia. They will have a minimum 18-month predoctoral contract with Himikode.

Specific Requirements for Candidates

• Be a national of a member state of the European Union.
• Non-EU nationals may apply as long as they have the necessary permits to stay in the country.
• Be enrolled or meet the requirements to enroll in an official doctoral program at a university within the SUG (Galician University System) in accordance with applicable academic regulations for the 2023/2024 or 2024/2025 academic year.
• The completion date of studies must be within the last five years, with certain exceptions where it may be on or after January 1, 2017.
• Not hold a Ph.D. degree.

Abstract of the Thesis Project

The thesis to be developed is integrated into the project of developing a new IoT (Internet of Things) device that evolves from an already developed algorithmic solution to a new range of solutions based on neural networks, machine learning, etc.

The specific project will involve researching and developing a technology platform for detecting and classifying machinery and equipment states and events in the HORECA sector (Hotels, Restaurants, and Cafés) based on AI.

The platform to be developed will consist of:

• Redesigning the electronic sensors of the IoT device to equip it with multispectral acoustic sensors;
• Designing and implementing the most suitable type and topology of neural network for multispectral sound event classification;
• Conducting fieldwork to capture physical signals from equipment in the HORECA sector;
• Generating synthetic signals from the physical signals captured to thoroughly train neural networks for all possible scenarios;
• Training the selected networks with original and synthetic signals for the HORECA scenarios based on their precision and performance;
• Adapting the neural network code to the limited requirements of the microprocessors used in the IoT device, as well as programming it;
• Redesigning the MQTT frames sent by the IoT device to the signal processors;
• Programming the unitary processing applications of MQTT measures in Himikode's back-end;
• Programming the statistical processing of MQTT measurements acquired from all HORECAs to produce actionable information for clients.

In summary, the challenge posed to the doctoral student is a vertical project where they can apply their knowledge in IoT, AI, and Big Data, with a strong focus on designing solutions based on neural networks and programming data infrastructure solutions.

The doctoral student will not start from scratch; Himikode has already successfully developed an IoT platform and has electronics specialists to accompany them during the sensor redesign phase. The responsibility of the doctoral student in this phase is to create the AI methodology that validates the most suitable sensors and iterates with the Electronics team to find the optimal result.

Reprogramming the IoT devices involves adapting the size of the designed neural networks to the limited RAM and CPU capacity of the microcontrollers used, for which the doctoral student must demonstrate competence in low-level languages such as C and C++.

Regarding the selection of suitable neural networks, the doctoral student will lead the interaction with the Gradiant technology center, with which Himikode is currently developing a similar project in the field of AI.

Finally, concerning the programming of systems and related infrastructure, Himikode already has a system for sending and storing MQTT frames and systems that process the measurements both individually and statistically. The doctoral student's task will be to evolve this infrastructure to adapt it to the new MQTT measures created and to develop infrastructure modules that appropriately post-process this information in suitable languages such as Go, Java, C++, etc.

Specific Knowledge

Programming skills:

• Primary Languages: Python, Java or Scala, C++ or C, Go
• Complementary Languages: PHP, SQL, R, MATLAB or Octave

Tools and Libraries: TensorFlow, Keras, PyTorch, Pandas, NumPy, SciPy, Tidyverse, Git

Additional Knowledge: Applied Mathematics, Machine Learning, Big Data

Other Valuable Profile Competencies

• Relevant Academic Experience: Having developed a Master's or Bachelor's thesis focused on neural networks, AI, or Big Data.
• Professional Experience: Experience in technology projects, machine learning, Big Data.

Additional Competencies:

• Ability to communicate technical results
• Initiative and ability to work independently
• Participation in relevant courses or certifications

Contact

If you are interested, send us your CV indicating in the subject Doutoramento Industrial Himikode to:

• citius.kmt@usc.es
• info@himikode.com