APINTECH LTD announces its participation in TRUST AI (www.trust.ai), a project budgeted at €4M and funded by the H2020 – Framework Programme for Research and Innovation and in particular by the FET (Future Emerging Technologies) instrument (Grant agreement: 952060). The project has an expected duration of four years and is co-ordinated by INESC TEC (Portugal). The consortium also includes TARTU ULIKOOL (EE), INSTITUT NATIONAL DE RECHERCHE ENINFORMATIQUE ET AUTOMATIQUE (FR), STICHTING NEDERLANDSE WETENSCHAPPELIJK ONDERZOEK INSTITUTEN (NL), APPLIED INDUSTRIAL TECHNOLOGIES (CY), LTPLABS LDA (PT), TAZI BILISIM TEKNOLOJILERI A.S. (TR)
About TRUST AI
Artificial Intelligence (AI) is a game-changer for a variety of sectors. AI’s boost to the global economy by 2030 is forecasted to be of $15.7 trillion (source PwC), and by 2021 80% of emerging technologies will have AI foundations (source Gartner). However, this massive potential will only be realized if humans, whether it is a doctor, a manager or a scientist, are able to trust the outcome of AI models, and use their own expertise to validate and improve them. This is virtually impossible with the current black-box models that dominate AI. These models achieve great performance but cannot be explained. The next generation of AI will therefore have to increase this explainability.
What will be developed?
Our approach promotes AI and humans working in concert to find better solutions, models that are both effective as well as comprehensible. Our tool is TRUST ⎼ a Transparent, Reliable and Unbiased Smart Tool. By being explainable-by-design, it will be transparent, reliable and able to prevent undesirable biases. Most of all, TRUST will allow humans to gain insight and understand the workings of the AI models.
How will this be done?
TRUST will be demonstrated and validated in three use cases: one in healthcare, for tumor treatment, one in online retail, for selecting delivery time slots, and one in energy, for supporting demand forecasts at buildings and beyond. Yet, the true potential of TRUST will go well beyond these trials, it will impact society at large and find applications in a wide range of sectors, such as banking, insurance, manufacturing and public administration.
TRUST AI in Energy
The energy related thread of activities in TRUST AI will seek to introduce and make use of xAI concepts for use cases such as building energy (and beyond) demand forecasting, and demand response. Critical in the overall approach is to transcend the sheer black box prediction and provide for stakeholder, guided decision support.