Tuesday, February 5, 2019

B4 Project

      I.        Introduction
a.   Define Artificial Intelligence/Machine Learning/Deep Learning
b.   Define Smart Infrastructure
c.    State the Argument (there will be adverse effects from using Artificial Intelligence which stem from the demographic of individuals in engineering and construction practices, thus encouraging counterproductive results to the betterment of society and future buildings we construct)
    II.        Advances in Current Engineering/Construction Practice
a.   Examples:
                                               i.     Drones or Unmanned Aerial Vehicles
                                             ii.     Sam 100 (Bricklaying Robot) or Construction Robots
                                            iii.     Comfy Application or HVAC Robots
                                            iv.     BIM or BLM Software
                                             v.     Chatbots
b.   Personal Comments
   III.        Future of Engineering/Construction Practice
a.   Examples
                                               i.     Based on Existing Technology i.e. Part II and Future Demand i.e. References
  IV.        Discussion of Artificial Intelligence and Connection to Human Behavior
a.   Define Characteristics of Machine Learning
                                               i.     Methods of Learning
                                             ii.     What Makes a ‘Good’ Machine?
b.   How does Human Behavior Impact Machine Learning
                                               i.     Before and During Implementation of Technology in Actual Situations
    V.        Psychological Implications of Using Artificial Intelligence Generally
a.   Human Dependency
                                               i.     Laziness
                                             ii.     Carelessness
                                            iii.     Lack of Understanding for Traditional Ways/Threat to Jobs
                                            iv.     State Some Positive Impacts for Fair Argument
b.   Social Views
                                               i.     Cultural Differences i.e. race, income, religion, politics, etc.
  VI.        Psychological Implications Inferred from Using Artificial Intelligence in Engineering/Construction
a.   Based on Part V and Psychology i.e. References
                                               i.     Emphasize Machine Learning
 VII.        Final Thoughts and Expert Opinions 
a.   Example Critics and their Arguments both for and against Part I.c.
VIII.        Conclusion
  IX.        References

My project will discuss how the use of artificial intelligence combined with human behavior can lead to negative repercussions in the future. Specifically, I would like to use machine learning, based on how it is used today and presumably in the future, as an example of the interconnectivity between machines and human thought and behavior to show that machines are not completely independent of social prejudices.

I chose this topic based on an article that I read about a self-driving Uber. One night a passenger was being transported by a self-driving Uber, and after each pedestrian walked by the Uber would identify them to the passenger. The car was unable to identify and nearly hit two African American men because it could not see nor recognize dark skin. This made me realize that even though machines can be adaptive to their surroundings, they are still influenced by the experiences that they have during the time of their construction. If there is not a lot of diversity in the groups of people who design or use artificial intelligences like machine learning then in essence, these machines will make seemingly unbiased judgments based on biased/limited interactions.

This is relevant to Intelligent Buildings because like the automotive industry, the engineering and construction industry is slowly adapting to incorporate advanced technologies to replace traditional human labor. I think the diversity of these machines in solving problems and interpreting the external environment is important to the overall success of accomplishing a construction task (ex: laying bricks, pouring cement, carrying material). This is why it is necessary to expose these machines to a multitude of foreign environments and people before actual implementation, as well as to monitor how these machines respond to problems over time and if this response is in alignment with the goals for having designed the machine.

Specific to Intelligent Buildings machines will adapt to do tasks like assist with the labor in building construction, automatically adjust interior conditions based on occupancy and weather, and monitor the deterioration of building components. A diversified machine will have been exposed to a number of building failures and have a better likelihood of solving any problems that arise during the buildings lifetime (especially if the same machines were used during the construction of the building).

The challenges with this topic are that much of the argument will stem from interpretive analysis. That is, I will be drawing conclusions from psychology and sociology experts as well as engineering and construction experts because there are not a lot of scientific resources surrounding this topic. There are arguably many benefits to artificial intelligence and this paper isn’t intended to detract from those; the intention for the paper is to bring awareness to a potential future problem that should be prepared for in order for it to be mitigated and ideally, but not likely, prevented.

Comments:

Alkiviadis – I appreciate that you are not only writing about building security through new technology but are also implementing a concrete deliverable. I would suggest that you expand further by learning but also predicting future technologies that could arise to solve building design issues in say your model for example.

Borelli – I like your topic and the fact that you are pulling from existing examples to explain the current capabilities in these machines. I would compare the traditional method of Energy/Life Cycle Analysis with that of using AI to see if there is, if any difference in the quality of the results.

YIDI LI – Your topics stems from much of what we learned in class. I would expand on the impact of urbanization both in America and in underdeveloped or building countries and draw conclusions on how the machines could be designed to address the concerns associated with those areas.

References:

No comments:

Post a Comment

Note: Only a member of this blog may post a comment.