Lab Report Analysis: Experimenting Artificial Intelligence 

Muhtasim Mushfiq 

Professor Crystal Rodwell 

ENGL 21007

City College of New York

Abstract

This paper compares and contrasts three lab reports that experiment using artificial intelligence to improve their initial experiments. “Blending Artificial Intelligence into PID Controller Design: A Biomedical Engineering Experiment” was about persuading students to use the particle swarm optimization (PSO) algorithm to make the results gathered from the IAE based tuning rules more advanced. (Oliveira & Cunha 2016). This lab report would be referred to as Lab 1. “A strength prediction model using artificial intelligence for recycling waste tailings as cemented paste backfill (CPB),” this experiment was about creating a prediction method for CPB and they used the boosted regression trees (BRT) algorithm for modeling non-linear relationships and also used the PSO algorithm for the BRT hyper-parameters tuning (Qi et al., 2018). This lab report would be referred to as Lab 2. The last lab report, “A Nearest Neighbour extension to project duration forecasting with Artificial Intelligence.” was about trying to use artificial intelligence to reduce the training set and limit the project’s duration and budget. This last lab report did not use the PSO or BRT algorithms like the Lab 1 and Lab 2, they used other methods such as, Planned Value(PV), Earned Duration(ED), and Earned Schedule(ES) (Wauters & Vanhoucke, 2016). These lab reports are compared through the use of the Writer’s Checklist on page 531 of Technical Communication by Mike Markel as a foundation to analyze its effectiveness (Markel, 2018).

      The first section being analyzed of a lab report will be the title. The title for all three lab reports was lengthy and specific. The titles conveyed the major focus of the study. Lab 1 used keywords such as artificial intelligence, PID control, and Biomedical Engineering. The keywords Lab 2 used were artificial intelligence, cemented paste backfill, and recycling waste tailings. The keywords Lab 3 used were artificial intelligence, Nearest Neighbour, and forecasting. The audience was different for some of the reports. In Lab 1, the intended audience was students and in Lab 2 and 3, the intended audience was professional engineers/scientists, so the keywords they used in the titles were adequate. Moving on to the abstract section, the abstract in all three lab reports clearly stated the problems that came forth during the experiments. The authors quickly summarized their approach. However, in Lab 1, the authors do not go into the key results that they have discovered as opposed to Lab 2 and 3 that summarize the results and conclusion with a sufficient amount. Next, the introduction section. The introduction of all three lab reports were very similar. They successfully gave background information that led to their research and the purpose of the experiments. Lab 2 briefly describes the methods while Lab 1 and Lab 3 give an outline at the end of the introduction of what is to come in the rest of the reports. Lab 1 was missing some portions of the abstract. Besides that, all three lab reports followed most of the requirements listed on the checklist for the title, abstract, and introduction. 

      The next section of the lab reports to be analyzed is the material and methodology section. In the materials section, the authors give a well-written description of each significant material that was used in each experiment. In the methodology section, the authors explain the procedure with great detail for the audience to understand exactly how they went along doing their experiments. As they described the procedure, they explained the formulas they used and presented valid measurements and calculations. Also, they used graphs, tables, and pictures as they went along to further explain the procedure. The authors also presented some problems they have faced in the experiment which can be seen in the graphs. And for those problems, they introduced a solution. In Lab 1, they changed their method of only using the PSO algorithm to start using the degree of freedom with lead-lag configurations (2DOF). The information they presented were in chronological order to be able to follow along without jumping around. 

      In the results section of the lab reports, all lab reports were successful in summarizing the data they have collected and explained it to a great amount. They made sure to emphasize some of the data through the graphs they presented. The graphs supported their claims and data. The authors refrained from using data that was not significant in the experiment. The authors were able to avoid interpreting and analyzing the data, they left that for the discussion portion of the lab report. In the discussion section, the authors discussed the hypothesis they addressed in the introduction. In Lab 1, as stated before, they believed the PSO algorithm would be best but it turned out that the degree of freedom with lead-lag configurations (2DOF) had the best performance. The authors of Lab 2 stuck with their initial claims and got the data they needed from the BRT algorithm and PSO-BRT algorithms. The authors successfully gave plausible explanations for the results. They also addressed major trends, values, and statistics which are also shown in the graphs presented. The authors, also, related their work to other experiments that have already occurred. They were able to support their claims with the data and results they received. 

      In the conclusion section of the lab reports, all lab reports summarized the entire experiment. They reviewed the purpose of the experiments and gave a recap of important findings throughout the experiment. The authors did not introduce new topics in the conclusion, however, they did explain what these findings could lead to. Lab 1 did not have an appendix section while Lab 2 and 3 did have an appendix section. The information Lab 2 and Lab 3 presented in the Appendix did fit the data and experiment which means that no unnecessary data was given. Lab 2 and 3 also had an acknowledgment section where they thanked those who helped contribute to the experiment. Finally, all of the lab reports had a reference section where they cited any sources they have used. The citations contained complete and accurate information. 

As can be seen, the lab reports were identical in most of the sections when it comes to the structure/format. Lab 1 had the most differences from Lab 2 and Lab 3 because it did not have many of the sections they had. Also, Lab 1 found a better method of conducting the experiment while in the middle of the experiment. However, all lab reports met the majority of the requirements a lab report should meet by following the checklist on the Writer’s Checklist. 

References

Cunha, J.B., Oliveira P.B. (2016). Blending Artificial Intelligence into PID Controller Design: A Biomedical Engineering Experiment. Science Direct, 49(6), 366-371. Retrieved from https://www.sciencedirect.com/science/article/pii/S2405896316304116.

Markel, M., & Selber, S. A. (2018). Technical communication. Boston: Bedford/St. Martins.

Qi, C., Fourie, A., Chen, Q., Zhang, Q. (2018, May 10). A strength prediction model using artificial intelligence for recycling waste tailings as cemented paste backfill. Science Direct, 183, 566-578. Retrieved from https://www-sciencedirect-com.ccny-proxy1.libr.ccny.cuny.edu/science/article/pii/S0959652618304670#sec5.

Wauters, M., Vanhoucke, M. (2017, June 16). A Nearest Neighbour extension to project duration forecasting with Artificial Intelligence. Science Direct, 259(3), 1097-1111. Retrieved from https://www-sciencedirect-com.ccny-proxy1.libr.ccny.cuny.edu/science/article/pii/S0377221716309432.