2011 Scientific Research and Experimental Development tax credit claim
T661 - Part 2 – Project Information
Section B – Experimental development
240 - What technological advancements were you trying to achieve? (Maximum 350 words)
General purpose learning and thinking software for control of robots based on artificial neural networks (ANN).
- Advancements in the architecture of a hierarchical structure for ANN nodes.
- The use of binary neurons (binons) with reinforcement learning.
- Network growth from the addition of nodes for parallel and sequential pattern recognition.
- Conversion of graduated sensor readings into symbolic stimuli.
- Transfer of short term memory traces into long term memory.
- Learnt action habits integrated into an associated hierarchical behavior network.
- Parallel execution of subconscious action habits.
242 - What technological obstacles/uncertainties did you have to overcome to achieve those advancements? (Maximum 350 words)
Current robot control software is not general purpose. It is specifically developed to function in a predefined environment with predefined tasks to be performed. Learning capability, if present, is limited to this environment. General purpose artificial intelligence architectures exist but are not designed for robot control i.e. they are not grounded on sensory input and device activation. Examples include ACT-R, SOAR, and HTM. DRAMA, MAXSON and Brook’s COG are experimental architectures designed for robot control but have not been applied with any success. This research is developing a new architecture to solve the problem. However there are no hierarchical ANNs based on binons that are feed-forward and use reinforcement learning. There are also no ANNs that grow nodes in a hierarchy for each new object or sequence recognized or for each new action habit learnt.
244 – What work did you perform in the tax year to overcome the technological obstacles/uncertainties described in Line 242? (Summarize the systematic investigation) (Maximum 700 words)
Approach: Increase the complexity of the ANN structure and algorithms to process the requirements at greater complexity while still handling the lower complexity features. Run regression tests on already working lower complexity features. Determine its success at learning and thinking based on the observation of its actions in artificial test environments simulated in software and by inspection of its internal memory traces and processes.
Areas of research performed this year - Two pieces of software are being crafted;
A) In the area of object and sequence recognition from multi-sensor senses producing graduated and discrete (symbolic) stimulus readings using a hierarchy of binons. Each new version of this software is then integrated into B.
B) In the area of short term and long term memory, learning, thinking and a hierarchical action habit structure.
In area A:
- Evolved the tree structure representation and modified the approach for identifying objects and sequences of stimuli. Four versions developed.
- Changed recognition from sequences of parallel stimuli to parallel combinations of short term memory sequences.
- Restructured the combination of lowest level features in the hierarchy.
- Restructured the “Where” and the “What” information for identifying an object.
- Combined the “Where” information with the “What” information at all levels of the recognition hierarchy rather than just at the top.
In area B:
- Integrated the pattern recognition algorithms from the A software into the B software. This was done with the four different pattern recognition approaches.
- Separated the change of a stimulus that attracts attention (makes it conscious) from the interest level that it is given when it becomes conscious.
- Used the concept of dependent and independent stimuli to determine when the combinations of parallel stimuli become conscious.
- Removed the use of an explicit concentration level.
- Determined the interest in redoing an action habit based on the interest of the goal rather than the redo-interest of the habit from its most recent execution.
- Restructured the representation for action habits when they are active.
- Changed object recognition from parallel combinations of sequences to sequences of parallel stimuli to properly integrate with active action habits.
- Remodeled thinking as the mental repetition of the new action habit structure.