REACH startup Sensinov helps their clients achieve improvements in their energy performance, a cleaner environment, smarter buildings, and an overall improved quality of life for their users.
Hi Eliana! Can you present your team and company to us?
Hi REACH! Sensinov is a startup on its way to revolutionize the way buildings are managed.
We are a team of 16, with core competences and a solid track record on AI/ML, Energy Management, Cloud and Edge computing.
My name is Eliana, and I joined back in 2020 as an R&D lead. Our CEO and co-founder, Mahdi Ben Alaya, founded Sensinov in 2016 with a vision in mind: break down the technical and economic barriers that have kept buildings -and building managers- in the dark for so long.
This vision has carried us out through our company and product development over the years. We now successfully market a complete hardware and software “plug-and-play” solution – our Intelligent Building Management System.
When combined with smart retrofitting, IoT and state-of-the-art AI, the latter allows:
- to enforce global energy policies across entire building stocks with just a few clicks;
- easy data sharing, based on Semantic Interoperability, to unlock the full potential of buildings;
- quick and easy deployment, designed to fit any size of building and multiple use cases.
So far, we have helped our partners achieve lasting improvements on their energy performance (>30%), a cleaner environment, smarter buildings, and an overall improved quality of life for their users.
How did you learn about REACH Incubator and what made you apply?
We first learned about REACH through the Alliance for the Internet of Things Innovation (AIOTI) ecosystem. As a startup that is heavily invested in R&D and innovation, we try to be as active as possible in the initiatives that are shaping the future of IoT and Smart Buildings in Europe, such as AIOTI.
Our main motivator for applying was the challenge proposed by CERTH: we had already started developing AI-based solutions for detecting anomalies and preventing device failures and we were interested in taking those developments one step further, by building a continuous look-ahead Energy Optimization System.
The technical and mentoring resources that were proposed during all phases of the incubation were our definite push to join!
You are tackling CERTH’s challenge Occupancy-driven monitoring & multi-factor recommendation systems for Energy Efficiency In Buildings. How do you think experimenting with and solving their data value chain challenge will set the foundations for scaling your solution?
The challenge from CERTH is to build a non-intrusive engine, capable of inferring building occupancy based on data collected from the HVAC systems (e.g., energy measurements), as well as indoor/outdoor data points (e.g., temperature, humidity, etc.) and to develop optimal, occupancy-informed, heating strategies for HVACs.
Occupancy-based (or demand-based) strategies have been repeatedly proven in the literature to reduce energy consumption by up to 50%. However, these methods remain vastly underused. Why? common available monitoring methods (e.g., IoT-based or model based) fall flat when it comes to precisely inferring occupancy, while also quickly increasing the total cost of such deployments.
As a fully integrated solution, our platform will use a simple and flexible API to build machine learning powered recommendations for efficient energy management, driving down operational costs, and optimizing self-consumption.
Please present your solution and elaborate on how it differentiates from the competition.
PLATONIC, will develop an occupancy & renewable energy sources (RES)-aware Heating, Ventilation and Air Conditioning (HVAC) recommendation system for commercial buildings. We focus on HVAC systems since they are, on average, responsible for over 40% of a building’s energy consumption.
PLATONIC’s application architecture is composed of 3 asynchronous, modular, and reusable building-block components:
- The occupancy model, which is based on environmental signal deconvolution. This method allows us to exploit readily available HVAC data in our system, such as temperature, CO2 levels, and HVAC ventilation setpoints.
- The Renewable Energy Sources (RES) model, which uses Regression Analysis to identify the relation between weather-related variables and local energy production.
- And finally, the HVAC recommendation system model, which implements an innovative approach based on Deep Reinforcement Learning to provide a set of recommendations that optimize HVAC operations.
PLATONIC’s approach to efficient HVAC management will separate us from the competition by offering:
- An Open-Source based Framework that will accompany the growing demand for additional technological progress in this area, while ensuring user’s comfort and wellbeing.
- A more holistic view of the building, that includes local production of renewables to improve self-consumption of buildings, allowing for additional energy savings and the overall reduction of operational costs.
- Automatic adjustment of heating and air conditioning programs with the simple click of a button.
Do you foresee any obstacles in successfully developing and commercializing your solution?
Enabling smart autonomic buildings requires breaking down building systems into a network of building components, that rely on appropriate software to collect & analyze significant amounts of data, enabling physical buildings to be managed, quickly understood & reconfigured as needed.
There are however some technical and technology-related risks involved. Even if the technology is fully developed nowadays (e.g., home automation, IoT, big data), autonomous buildings will require significant design and engineering efforts. AI plays a pivotal role in this effort, facilitating the operation of the collected data to reinforce the creation of high added value data chains, like the ones promoted by REACH.
We are confident that by combining our unique expertise and joining our efforts towards solving this challenge, we’ll develop a modular and performant solution that tackles some of the market’s biggest challenges.