About

Local AI is a technology startup founded in Kalamata, Greece in 2023 by young scientists, AI enthusiasts, developing innovative solutions for sustainability utilizing the power of AI for local and global users, such as, Municipalities, Regional Government, Energy Providers, Utility Companies, etc. Local AI was incorporated on February 21st, 2023, as an IKE and was registered to the relevant Chamber of Commerce with VAT no 802053224, ΓΕΜΗ no 16899034500. It is owned 100% by Georgios Vlachodimitropoulos. Local AI is focused on activities of bringing Artificial Intelligence Solutions in the service of Sustainability in collaboration with stakeholders such as Universities, Tech Hubs, Innovation Centers, etc.

Solution & Product

CNROpt - Charging Network Rollout Optimization Tool: AI enabled Geospatial Analytics tool for Automated Charging Network Expansions Planning. CNROpt is addressing needs of EV Charging Stations Network Operators, use AI to assist planning the network. As a result of CNROpt, we have developed in the MVP level a planning tool for Charging Station Networks, that uses AI to optimize the place and time of deployment of new Charging Stations. AI4CS project - AI for Charging Stations: AI4CS is using AI to guide their customers when they need to recharge their car. We built a mobile app that is used by EV owners, that uses AI to optimize when and where to charge, considering customer experience, Grid Load and Green Electricity Production. This project engaged us into EU-wide platform for energy grid data providing us insights on how to best guide the electricity consumers promoting Grid resilience and Green Electricity consumption. AI4FS project – AI for Forest Surveillance: We will build a Web based Geospatial Analytics platform that employs our AI algorithms processing real-time drone multispectral video footage, performing automated forest surveillance, monitoring thickness, health, events etc. EasyRide Green Urban Mobility: a navigation platform and mobile app addressed to citizens helping them make greener decisions for their everyday mobility. Space-enabled Environmental Dashboard: Geospatial AI extracting environmental information from satellite RS data.

Innovation & Scalability

We implement state-of-the-art timeseries forecasting AI models that have shown very high accuracy. We have trained Deep Learning Models to do Time Series Forecasting and we use LSTM and Transformers models as top performers: Transformers Model: Transformers constitute a class of Deep Learning Models that are suited to handle temporal dependencies on data. They are the backbone of modern Language Models (GPT)[4] and modern paradigms for Spatiotemporal Video Action Recognition. We developed our custom Transformer-based forecast model to perform the congestion-utilization forecast. LSTM Model: LSTM belong to that category of Recurrent neural networks have been successful in time series forecast, especially when seasonality and cycles need to be accounted for. We have leveraged previous works that employ different types of recurrent neural networks (LSTMs, Random Neural Networks, GRUs) for anomaly detection in IoT systems via time series forecasting for the problem of congestion/utilization forecast. These models currently perform excellent providing precise forecasts, outperforming the baseline ARIMA ML model. In addition we build state-of-the-art Geospatial Analytics Apps that provide powerful visualization to the end user.

Traction & achievements

We have started our operations seeking funding from the EU Horizon 2020 projects framework and we have been awarded 3 projects so far: 60k from the i-nergy (https://i-nergy.eu/) for the CNROpt project 100k from the Interconnect (https://interconnectproject.eu/) for the AI4CS project. 60k from the ImagineB-5G (https://imagineb5g.eu/) for the AI4FS project.

Team & Why Us

Our next steps include the collaboration with a major utility company and/or public organization to pilot our solutions in order to bring them in the TRL-7 production level. Our solutions target both the Private and the Public Sectors and our business strategy is pursuing a B2B2C model engaged with a major player in utilities (Energy, etc.) and a B2G model accordingly in association with Municipalities/Regional Administrations. For the B2B2C model we plan to provide AI solutions through the SaaS model enabling Utility companies upgrade their offerings to their subscribers with smart energy management tools to achieve both energy savings and reduce their Carbon Footprint. For the B2G model we will provide Municipalities/Regional Administrations with platforms to monitor the performance of their Green Transition projects, empowering citizens to make environmental friendly decisions for their mobility and energy consumption. Our company is founded by AI scientists recently graduated by leading Greek engineering school with graduate MSc and PhD from Rutgers University at USA. We offer specialized advanced AI algorithms such as Deep Learning, Generative AI, RCNN, etc. that can help various optimization tasks that are crucial in achieving lower Carbon Footprint in several domains, such as Energy, Mobility, etc.

Funding Needs

200k for AI4SB: AI for Smart Buildings. Use the experience and algorithms employed at the CNROpt and AI4CS projects to make a platform addressing Household Energy Management. Market it directly to consumers and/or in partnership with an energy Utility. 200k for Space-enabled Environmental Dashboard. Deploy a platform that can provide the Green Transition Performance Environmental metrics produced by applying AI models to the abundance of Satellite Remote Sensing. 100k for the EasyRide Green Urban Mobility navigation platform and mobile app addressed to citizens helping them make greener decisions for their everyday mobility.