A constant influx of information does not allow for a mind to enter a state of high neuroplasticity at a low-level cognitive load which results in a new problem for the brain to solve. If viewed from a Darwinian perspective, problem-solving falls to natural selection as evolutionary processes maintain the balance of chaos while the lack of functional by-products (such as insight) of the default mode network (DMN) are affected by reflective hinderance (Goldstein, 2018).
“A robot is a machine that carries out tasks or groups of tasks that have been pre-programmed in it. Robots are machines with additional features[,] hence more functionality. Robots are self-governing machines capable of making decisions without an external trigger.
A machine, on the other hand, has to be operated by a human to act” (Robots Science, 2020).
Reasoning and insight used in problem solving where priming and product interact to evaluate and employ reasonable resolutions, though not well understood as discussed by the Goldstein text (2018), in which divergent thinking remains a proficiency of humans. Despite having uncanny adaptative skills, collaboration with robotic systems demonstrates limitations that lead to risk, performance obstacles, and other threats to safe operations due to the cap in expert mental sets that create domain knowledge.
The collective work of human-robot corroboration faces limitations regarding effective communication. Clarity in recognition and comprehension of human actions by robots must be provided by direct programming and training. Due to the nature of internal versus external attention that promotes natural learning over time through a process of trial-and-error, the human reactions fail to meet anticipated scripts of robotic consoles, leading to the need for risk assessment in the workplace resulting from safety and performance obstacles that can be functionally related to error, regardless of source monitoring (Goldstein, 2018; Laura et al., 2017).
To construct a research project to examine how to demonstrate this aspect of cognition, the specific research question is:
“What features and absolute barriers for identifying limitations and boundary issues to reduce the limitations and errors of human-robotic system interactions using natural language could be implemented to address the potential source monitoring and mental set issues?”
Current projects such as the “Air-Cobot” in France which “aims to develop a collaborative mobile robot to obtain a human-robot inspection of an aircraft during maintenance operations before takeoff on an airport” offer positive feedback regarding interactions provided that the human operator supervises its mission, checks its non-destructive testing results, and intervenes if the robot is in trouble” (Futterlieb et. al., 2017).
To conduct an effective risk assessment, threats to the security of operation, obstacles to efficient performance, and potential opportunities in error reduction must be considered. As language is the basis of communication for the human-robot interaction, variables regarding semantic development and lexical priming as discussed by Goldstein (2018) require determination to produce an effective problem-solving response with an increased level of accuracy in evaluation.
The challenges facing researchers and developers range from the approach in foundational model development itself, as concepts range from domino models to fault trees offering potential for positive system integration to testing and implementation into a dynamic workplace (Guastello, 2013). In considering that the “lengthy trial-and-error process” of problem-solving is functionally relational to insight, the use of robotic systems in the workplace based on human cognition must overcome mental sets and preconceptions that hinder creativity as the process is iterated (Goldstein, p. 378, 2018).
In solving problems, reflecting on how the problem has been presented is often overlooked, inadvertently or subconsciously, due to external or internal stressors. By integrating lexical priming with the use of chunked patterning, the space created in the void of information could trigger neuroplastic movement to begin functionality of the default mode network (DMN) (Goldstein, 2018).
Lexical ambiguity is a situation where “words can often have more than one meaning” (Goldstein, p. 327, 2018). As a result, the disparity between the human interaction with a robotic system based on factorial algorithms and deep-learning modules that do not account for activities of the default mode network (DMN) as noted in current “available neuroimaging evidence, [which supports] the embodied semantics approach, although the complex organization of the human motor cortex imposes limits to the anatomical localization of complex actions” (Fernandino & Iacoboni, p. 53, 2010).
Artificial intelligence (AI) based on analytical algorithms employing a lexically primed framework for human operation would focus on non-insight problems that could potentially improve interactions for the human factor as “restructuring” would occur in the natural state to allow an efficient performance time with less “stress and anxiety” that lead to hazardous working conditions (Goldstein, pp. 357-358; Guastello, p. 235, 2013). Those based singularly on deep learning also fail to synthesize insight to create. If deep-learning integrated the property of insight based on analytical problem-solving processes, a new human-centric design could be developed that would create a more human-to-human simulation.
On the other hand, basing artificial intelligence (AI) on human cognition may not be the best course of action based on the system outcome requirements. It is possible that by utilizing human-animal interactions, a system could be developed based on established positive relationships that could be integrated to perform basic human functions but allow the human element to exist in the natural state of priming to product cycle. Examples include hummingbirds that would offer excellent insight for flight while Jindo dogs could offer characteristics of loyalty; elephants could be used as models for heavy lifting.
Another possibility in improving human centric designs would involve utilizing fMRI technology as discussed in Crawford’s presentation (2019), as it may be possible to create an interactive membrane for devices that would allow the charge of water to be used as an intermediary of communication through biosensors. Based on analogous thinking as discussed by the Goldstein text (2018), the idea of visual scanning results combined with the tracking of neuroplastic movements could be matched to readings in the changes of water specific to the appendage of interaction and placement in mental scanning.
The default mode network (DMN) processes with relation to internal and external sustained attention pose an area of pivotal research with specific applications to human-robotic system interactions. With research maintaining a primary focus on information processing, the development of human-machine interactions (HMI) has naturally followed sequence. Insight versus non-insight problems faced when interacting with machine programs or artificial intelligence (AI) due to human or machine error and limitation result in questions as to the validity of basing consolidated systems on the human cognitive structure (Goldstein, 2018; Guastello, 2013).
Despite this lack of foresight in design, human innovation has implemented discovery with the advent of the Internet and induction of human-machine interactions (HMI), though dangers in cognitive overload are not fully understood; such violations of the ability to think freely supported by Article 18 of the Universal Declaration of Human Rights (UNDP) which states that “everyone has the right to freedom of thought, conscience and religion” should be serious considerations for human factors professionals (United Nations, 2022). Future systems may benefit from the implementation of analogous analytics to combat cognitive overload while elevating human centric designs in workplace applications to ensure safety and error limitation in ventures of performance enhancement.
Crawford, M. (2019). Cognitive Neuroscience (Part One). Retrieved December 10, 2022, from Cognitive Neuroscience (Part One).
Fernandino, L. & Iacoboni, M. (2010). Are cortical motor maps based on body parts or coordinated actions? Implications for embodied semantics. (Vol. 112, Issue 1). 44-53. Brain and Language. Retrieved December 10, 2022, from https://doi.org/10.1016/j.bandl.2009.02.003.
Futterlieb, M., Frejaville, J., Donadio, F., Devy, M., & Larnier, S. (1970, January 1). [PDF] Air-Cobot: Aircraft enhanced inspection by smart and collaborative robot: Semantic scholar. [PDF] Air-Cobot: Aircraft Enhanced Inspection by Smart and Collaborative Robot | Semantic Scholar. Retrieved December 10, 2022, from https://www.semanticscholar.org/paper/Air-Cobot-%3A-Aircraft-Enhanced-Inspection-by-Smart-Futterlieb-Frejaville/0a4c4c9a56465b700e4e0d68f683b8a344cd02f9
Goldstein, E. B. (2018). Cognitive psychology: connecting mind research and everyday experience (5th ed.). Wadsworth Cengage Learning.
Guastello, S. J. (2013). Human factors engineering and ergonomics: A systems approach, second edition. Taylor & Francis Group.
Laura, M. H., Narber, C., Bekele, E., Sangeet, S. K., & J, G. T. (2017). Human modeling for human–robot collaboration. The International Journal of Robotics Research, 36(5-7), 580-596. 10.1177/0278364917690592
Robots Science. (2020). Robot vs machine: Difference between a robot and a machine? Robots Science. Retrieved December 10, 2022, from https://www.robotsscience.com/resource/difference-between-a-robot-and-a-machine/#:~:text=Robots%20are%20self%2Dgoverning%20machines,like%20turning%20on%20an%20umbrella.
United Nations. (2022). Universal Declaration of Human Rights. United Nations. Retrieved December 10, 2022, from https://www.un.org/en/about-us/universal-declaration-of-human-rights