We have developed an architecture made with an algorithmic scaffolding that hosts trainable neural network modules. It follows the affordance competition hypothesis. There is one loop that detects affordable actions. The output is a “salience” function in the control space. This action-priming loop uses the previously trained inhibitory and excitatory modules (for example, a module that produces obstacle-inhibition maps). The agent can improve its behaviour by learning new sensorimotor loops (action discovery) and by improving the mapping from sensory to salience maps of the existing loops. Action-selection follows via a robust algorithm (analogous to the human basal ganglia), which works by competition among affordances. The choices of the agent are explainable (the individual affordances prior to and after selection are visible). Furthermore, the agent can learn to navigate the affordance landscape via biasing action selection and mirror human intentions via a similar biasing mechanism. One notable aspect of this agent is its ability to produce emergent (logical) behaviours.
Addressing target audiences and expressing needs
- Business partners – SMEs, Entrepreneurs, Large Corporations
- I/we wish to transfer my/our IPR to an interested party
- Venture Capital
Our objective is to transfer the technology to the market. We have different options to follow:
- we are looking for medium-large enterprises for licensing and collaboration with the aim to transfer the technology on an industrial use case;
- we are looking for public and private investors for the development of a vehicle prototype and increase the TRL from 6 to 7;
- we are looking for public and private investors for the development of the cloud dreaming technology and increase the TRL from 6 to 7-9;
- we are looking for Venture Capital to invest in a startup that has a clear Business Model.
- Public or private funding institutions
- Other Actors who can help us fulfil our market potential
- Private Investors
R&D, Technology and Innovation aspects
The current technology readiness level of the driving agent is 6 since the technology has been demonstrated in the context of the automotive environment. A vehicle has been equipped with the driving agent, that has been tested with short tasks on a private track. The objective is to demonstrate the technology at higher TRL levels. We have different options to follow:
- we are looking for medium-large enterprises for collaboration and validate the technology in an industrial use case;
- we are looking for public and private investors for the development of a vehicle prototype and increase the TRL from 6 to 7;
- we are looking for public and private investors for the development of the cloud dreaming technology and increase the TRL from 6 to 7-9;
- we are looking for Venture Capital to invest in a startup that can start marketing the solution.
In the case of start-up creation, the business model based on a hybrid software solution. It is a combination of on-premises and SaaS models which lets customers have the driving agent applications installed on vehicles and take advantage of cloud services as well. The balance of the strategies is determined by customers according to their needs. The hybrid model permits to transfer of data between cloud services and on-vehicle control units; allows sending huge files and making seamless updates; allows storing sensitive data on-vehicle.
Social Impact: autonomous cars have several benefits –> Safety, most of these accidents (over 90%) are caused by humans, saving over a million lives each year, significant injuries reductions and its related healthcare treatments. Reducing these accidents could result in massive savings and better use of healthcare facilities. The technology will permit also access to cars usage to disabled and old people helping them to be more autonomous. New job creation due to the increase of the related economic impacts.
Environmental Impact: it is believed that reductions in Greenhouse Gas Emissions emissions through autonomous vehicles adoption will be supported by more efficient driving and possible increases in new car smart mobility solutions.
Economic Impact: a new report predicts a $7 trillion self-driving economy will generate new markets such as the car data industry. Other areas that will benefit is the IT business, road infrastructure, mechanics for reparation.
