scholarly journals A cognitive robot equipped with autonomous tool innovation expertise

2020 ◽  
Vol 10 (2) ◽  
pp. 2200
Author(s):  
Handy Wicaksono ◽  
Claude Sammut
Keyword(s):  
3D Printing ◽  
Tool Use ◽  
General Model ◽  
Search Space ◽  
Complex Task ◽  
Trial And Error ◽  
Constraint Solver ◽  
Action Model

Like a human, a robot may benefit from being able to use a tool to solve a complex task. When an appropriate tool is not available, a very useful ability for a robot is to create a novel one based on its experience. With the advent of inexpensive 3D printing, it is now possible to give robots such an ability, at least to create simple tools. We proposed a method for learning how to use an object as a tool and, if needed, to design and construct a new tool. The robot began by learning an action model of tool use for a PDDL planner by observing a trainer. It then refined the model by learning by trial and error. Tool creation consisted of generalising an existing tool model and generating a novel tool by instantiating the general model. Further learning by experimentation was performed. Reducing the search space of potentially useful tools could be achieved by providing a tool ontology. We then used a constraint solver to obtain numerical parameters from abstract descriptions and use them for a ready-to-print design. We evaluated our system using a simulated and a real Baxter robot in two cases: hook and wedge. We found that our system performs tool creation successfully.

scholarly journals An Experiment on the Use of Genetic Algorithms for Topology Selection in Deep Learning

2019 ◽  
Vol 2019 ◽  
pp. 1-12 ◽  
Author(s):  
Fernando Mattioli ◽  
Daniel Caetano ◽  
Alexandre Cardoso ◽  
Eduardo Naves ◽  
Edgard Lamounier
Keyword(s):  
Genetic Algorithms ◽  
Deep Learning ◽  
Deep Neural Networks ◽  
State Of The Art ◽  
Complex Task ◽  
Trial And Error ◽  
Complex Solution ◽  
Computation Algorithms ◽  
Computational Resources ◽  
Topology Selection

The choice of a good topology for a deep neural network is a complex task, essential for any deep learning project. This task normally demands knowledge from previous experience, as the higher amount of required computational resources makes trial and error approaches prohibitive. Evolutionary computation algorithms have shown success in many domains, by guiding the exploration of complex solution spaces in the direction of the best solutions, with minimal human intervention. In this sense, this work presents the use of genetic algorithms in deep neural networks topology selection. The evaluated algorithms were able to find competitive topologies while spending less computational resources when compared to state-of-the-art methods.

scholarly journals Specialization in the vicarious learning of novel arbitrary sequences in humans but not orangutans

2020 ◽  
Vol 375 (1805) ◽  
pp. 20190442 ◽  
Author(s):  
Elizabeth Renner ◽  
Eric M. Patterson ◽  
Francys Subiaul
Keyword(s):  
Tool Use ◽  
Comparative Approach ◽  
Cultural Learning ◽  
Vicarious Learning ◽  
Trial And Error ◽  
Older Children ◽  
Learning Conditions ◽  
Error Learning ◽  
Cognitive Feature ◽  
Do So

Sequence learning underlies many uniquely human behaviours, from complex tool use to language and ritual. To understand whether this fundamental cognitive feature is uniquely derived in humans requires a comparative approach. We propose that the vicarious (but not individual) learning of novel arbitrary sequences represents a human cognitive specialization. To test this hypothesis, we compared the abilities of human children aged 3–5 years and orangutans to learn different types of arbitrary sequences (item-based and spatial-based). Sequences could be learned individually (by trial and error) or vicariously from a human (social) demonstrator or a computer (ghost control). We found that both children and orangutans recalled both types of sequence following trial-and-error learning; older children also learned both types of sequence following social and ghost demonstrations. Orangutans' success individually learning arbitrary sequences shows that their failure to do so in some vicarious learning conditions is not owing to general representational problems. These results provide new insights into some of the most persistent discontinuities observed between humans and other great apes in terms of complex tool use, language and ritual, all of which involve the cultural learning of novel arbitrary sequences. This article is part of the theme issue ‘Ritual renaissance: new insights into the most human of behaviours’.

Compensating the cell-induced light scattering effect in light-based bioprinting using deep learning

2021 ◽  
Author(s):  
Jiaao Guan ◽  
Shangting You ◽  
Yi Xiang ◽  
Jacob Schimelman ◽  
Jeffrey Alido ◽  
...  
Keyword(s):  
Deep Learning ◽  
Light Scattering ◽  
3D Printing ◽  
Data Augmentation ◽  
Parameters Optimization ◽  
Training Data ◽  
Trial And Error ◽  
Scattering Effect ◽  
Scattering Effects ◽  
Light Scattering Effect

Abstract Digital light processing (DLP)-based 3D printing technology has the advantages of speed and precision comparing with other 3D printing technologies like extrusion-based 3D printing. Therefore, it is a promising biomaterial fabrication technique for tissue engineering and regenerative medicine. When printing cell-laden biomaterials, one challenge of DLP-based bioprinting is the light scattering effect of the cells in the bioink, and therefore induce unpredictable effects on the photopolymerization process. In consequence, the DLP-based bioprinting requires extra trial-and-error efforts for parameters optimization for each specific printable structure to compensate the scattering effects induced by cells, which is often difficult and time-consuming for a machine operator. Such trial-and-error style optimization for each different structure is also very wasteful for those expensive biomaterials and cell lines. Here, we use machine learning to learn from a few trial sample printings and automatically provide printer the optimal parameters to compensate the cell-induced scattering effects. We employ a deep learning method with a learning-based data augmentation which only requires a small amount of training data. After learning from the data, the algorithm can automatically generate the printer parameters to compensate the scattering effects. Our method shows strong improvement in the intra-layer printing resolution for bioprinting, which can be further extended to solve the light scattering problems in multilayer 3D bioprinting processes.

Developing and implementing 3D printing services in an academic library

2015 ◽  
Vol 33 (4) ◽  
pp. 472-479 ◽  
Author(s):  
Gillian Andrea Nowlan
Keyword(s):  
3D Printing ◽  
Best Practices ◽  
Policy Development ◽  
New Technology ◽  
Academic Library ◽  
Pilot Project ◽  
Trial And Error ◽  
Content Type ◽  
3D Objects ◽  
The University

Purpose – The purpose of this paper is to describe the development of a 3D printing pilot project and 3D printing library service. Policy development, instruction, and best practices will be shared and explored. Design/methodology/approach – This paper describes the implementation of 3D printing at the University of Regina Library and details successes, failures, and modifications made to better provide 3D printing services. This paper outlines one academic library’s experience and solutions to offering 3D printing for university patrons. Findings – Although 3D printing has been around for a while, it still requires trial and error and experience in order to print successfully. Training and instruction is needed to run the 3D printer and understand how to develop 3D objects that will print successfully. Originality/value – There have been many publications on 3D printing, but few that discuss problem solving, best practices, and policy development. 3D printing provides a way for patrons to learn about new technology and use that technology to help support learning.

Coas: Combined Object-Action Selection: A Human Factors Experiment

1992 ◽  
Vol 36 (4) ◽  
pp. 316-320 ◽  
Author(s):  
J. F. Kelley ◽  
J. Ukelson
Keyword(s):  
Task Performance ◽  
Action Selection ◽  
Complex Task ◽  
Interaction Techniques ◽  
Simple Task ◽  
File Management ◽  
Standard Object ◽  
Management Application ◽  
Action Model ◽  
High Level

12 participants with a high level of domain experience used two different, mouse-based, interaction techniques to carry out three workstation file management tasks of varying complexity. One technique followed a standard Object-Action model; the other was a newly developed technique called COAS (Combined Object-Action Selection). There was little difference in performance on a simple task; performance for participants using the new technique was 38% faster on a moderately complex task and was 21% faster on a complex task. The file management application, interaction techniques and experiment were implemented in an OS/2 Presentation Manager style using ITS (Interactive Transaction System).

A REAL-TIME SCHEDULING ALGORITHM FOR TASKS WITH RESOURCE CONTENTION ON A MULTIPROCESSOR

1994 ◽  
Vol 04 (03) ◽  
pp. 243-253
Author(s):  
DAL-SOO RYANG ◽  
KYU HO PARK
Keyword(s):  
Real Time ◽  
General Model ◽  
Resource Constraints ◽  
Scheduling Algorithm ◽  
Search Space ◽  
Real Time Scheduling ◽  
Multiple Processors ◽  
Time Scheduling ◽  
Two Measures ◽  
The Impact

Our scheduling algorithm is based on a general model with timing and resource constraints which permits OR requests. In order to keep run-time costs low, we propose an algorithm that does not search the whole search space. This paper defines two measures, survivability and impact, for scheduling tasks conflicted for some resources. The survivability is a metric to show how urgent a task is, and how constrained it is by its resources. The impact of a resource for a task measures how much other tasks are influenced by the allocation of the resource to the task. Our scheduling algorithm uses the survivability to schedule tasks on multiple processors. After a task is picked out to be run in a time slice using the survivability, the least impact resources are allocated from several alternative resources.

scholarly journals Towards A Relational Approach For Tool Creation By Robots

2017 ◽  
Author(s):  
Handy Wicaksono
Keyword(s):  
3D Printing ◽  
Past Experience ◽  
Complex Task ◽  
Relational Approach

Like a human, a robot may benefit from being able to use a tool to solve a complex task. When an appropriate tool is not available, a very useful ability for a robot would be to create a novel one based on its past experience. With the advent of inexpensive 3D printing, it is now possible to give robots such an ability, at least to create simple tools. We propose a method for learning how to use an object as a tool and, if needed, to design and construct a new tool.

scholarly journals Optimizing Constraint Solving via Dynamic Programming

2019 ◽  
Author(s):  
Shu Lin ◽  
Na Meng ◽  
Wenxin Li
Keyword(s):  
Dynamic Programming ◽  
Optimization Problems ◽  
Optimal Solution ◽  
Search Space ◽  
Constraint Solving ◽  
Constraint Optimization ◽  
Objective Functions ◽  
Constraint Solver ◽  
Constraint Optimization Problems ◽  
Finite Domains

Constraint optimization problems (COP) on finite domains are typically solved via search. Many problems (e.g., 0-1 knapsack) involve redundant search, making a general constraint solver revisit the same subproblems again and again. Existing approaches use caching, symmetry breaking, subproblem dominance, or search with decomposition to prune the search space of constraint problems. In this paper we present a different approach--DPSolver--which uses dynamic programming (DP) to efficiently solve certain types of constraint optimization problems (COPs). Given a COP modeled with MiniZinc, DPSolver first analyzes the model to decide whether the problem is efficiently solvable with DP. If so, DPSolver refactors the constraints and objective functions to model the problem as a DP problem. Finally, DPSolver feeds the refactored model to Gecode--a widely used constraint solver--for the optimal solution. Our evaluation shows that DPSolver significantly improves the performance of constraint solving.

scholarly journals Shells as ‘extended architecture’: to escape isolation, social hermit crabs choose shells with the right external architecture

2020 ◽  
Vol 23 (6) ◽  
pp. 1177-1187
Author(s):  
Jakob Krieger ◽  
Marie K. Hörnig ◽  
Mark E. Laidre
Keyword(s):  
Experimental Design ◽  
3D Printing ◽  
Cognitive Abilities ◽  
Physical Space ◽  
Solitary Confinement ◽  
Hermit Crabs ◽  
Trial And Error ◽  
X Ray ◽  
The World ◽  
The Right

AbstractAnimals’ cognitive abilities can be tested by allowing them to choose between alternatives, with only one alternative offering the correct solution to a novel problem. Hermit crabs are evolutionarily specialized to navigate while carrying a shell, with alternative shells representing different forms of ‘extended architecture’, which effectively change the extent of physical space an individual occupies in the world. It is unknown whether individuals can choose such architecture to solve novel navigational problems. Here, we designed an experiment in which social hermit crabs (Coenobita compressus) had to choose between two alternative shells to solve a novel problem: escaping solitary confinement. Using X-ray microtomography and 3D-printing, we copied preferred shell types and then made artificial alterations to their inner or outer shell architecture, designing only some shells to have the correct architectural fit for escaping the opening of an isolated crab’s enclosure. In our ‘escape artist’ experimental design, crabs had to choose an otherwise less preferred shell, since only this shell had the right external architecture to allow the crab to free itself from isolation. Across multiple experiments, crabs were willing to forgo preferred shells and choose less preferred shells that enabled them to escape, suggesting these animals can solve novel navigational problems with extended architecture. Yet, it remains unclear if individuals solved this problem through trial-and-error or were aware of the deeper connection between escape and exterior shell architecture. Our experiments offer a foundation for further explorations of physical, social, and spatial cognition within the context of extended architecture.

A Hybrid GABFO Scheduling for Optimal Makespan in Computational Grid

2014 ◽  
Vol 5 (3) ◽  
pp. 57-83 ◽  
Author(s):  
Shiv Prakash ◽  
Deo Prakash Vidyarthi
Keyword(s):  
Optimization Problem ◽  
Hybrid Approach ◽  
Optimal Solution ◽  
Characteristic Parameter ◽  
Search Space ◽  
Computational Grid ◽  
Complex Task ◽  
Proposed Model ◽  
Evolutionary Technique ◽  
Complex Optimization

Scheduling in Computational Grid (CG) is an important but complex task. It is done to schedule the submitted jobs onto the nodes of the grid so that some characteristic parameter is optimized. Makespan of the job is an important parameter and most often scheduling is done to optimize makespan. Genetic Algorithm (GA) is a search procedure based on the evolutionary technique that is able to solve a class of complex optimization problem. However, GA takes longer to converge towards its near optimal solution. Bacteria Foraging Optimization (BFO), also derived from nature, is a technique to optimize a given function in a distributed manner. Due to limited availability of bacteria, BFO is not suitable to optimize the solution for the problem involving a large search space. Characteristics of both GA and BFO are combined so that their benefits can be reaped. The hybrid approach is referred to as Genetic Algorithms Bacteria Foraging Optimization (GABFO) algorithm. The proposed GABFO has been applied to optimize makespan of a given schedule in a computational grid. Results of the simulation, conducted to evaluate the performance of the proposed model, reveal the effectiveness of the proposed model.

Sign in / Sign up

Export Citation Format

Share Document