Tag: Howard Weiss

Guest Post: Artificial Intelligence and Public Transportation

Prof. Howard Weiss shares his insights with opur readers monthly.

AI is becoming an increasingly important tool in public transportation and traffic management. Traffic cameras have been used since 1956, when they were introduced in London. Modern AI-enabled cameras, however, do more than record images. They can interpret what they see and identify traffic violations and other events in real time.

Cities and transportation agencies worldwide are experimenting with this technology. In Greece, AI cameras identify drivers who fail to wear seat belts, who use cell phones, or who exceed speed limits. In Goa, India, 26 locations have installed cameras to identify similar violations. Mississippi uses AI cameras to alert officers to traffic violations. Houston is purchasing AI cameras to improve traffic flow. In the Dallas–Fort Worth area, AI cameras assist criminal investigations

AI cameras are also being used on buses. In Philadelphia, SEPTA uses AI-powered cameras to identify vehicles illegally parked and blocking buses. Broward and Miami-Dade Counties in Florida use cameras mounted on school buses to identify motorists who fail to stop when buses are loading or unloading students.

Benefits and Costs The primary benefit of AI cameras is improved public safety. Several communities have reported reductions in traffic violations after cameras were installed. AI can also reduce the number of police officers required for traffic enforcement and can speed citation processing. 

Citywide systems can require millions of dollars in capital investment, as well as expenses for software development and preventive and breakdown maintenance, as discussed in Chapter 17 of your Heizer/Render/Munson textbook. Fines may offset some costs.

Privacy is another concern. Although systems identify vehicles rather than drivers, they can collect information about where vehicles travel and when. This raises important questions about data retention, access, and appropriate use.

Implementation and Quality Control The experience of several communities demonstrates the importance of quality control. Miami-Dade and Philadelphia both used warning periods before imposing fines. Miami-Dade later suspended its program because of inaccurate citations and subsequently modified the system to improve consistency and transparency. In Greece, reports found significant discrepancies between AI-generated citations and violations ultimately determined to be valid. Some drivers were cited even though they had stopped because of an ambulance or police vehicle.

Classroom Discussion questions:

  1. What rules should governments establish regarding data collection, storage, and access?
  2. What quality-control procedures should be required before a driver is fined based on an AI-generated violation?

 

Guest Post: A Class Exercise Relating Productivity and the Olympics

HowardWeiss2Howard Weiss is Professor of Operations Management at Temple University. He has developed both POM for Windows and Excel OM for our text
When I teach productivity, I like to explain to the students that while output/input seems simple, it can be difficult to assess.  The Heizer/Render textbook includes multifactor productivity (Chapter 1) where the multiple factors are all inputs, and I like to ask my students to take it a step further.

There are many web sites that list the Olympic medal count for any Olympics. I show my students the top medal-winning countries from one of the more recent Olympics, such as those listed at http://en.wikipedia.org/wiki/2012_Summer_Olympics_medal_table.

When I am in the classroom I ask my students, “What information would you want in order to answer the question, “Which of the 5 countries shown below was the most productive in the 2012 Olympics?” This raises many interesting issues. Students will discuss the different inputs that might be used such as population, dollars spent training the Olympians and number of athletes competing. Some students will note that different outputs can be used. For example, outputs could be total medals or a weighted average of gold, silver and bronze medals. For the 2012 Olympics, I point out that based on the sorting in Wikipedia, it is using only gold measures, as its output measure as can be seen when comparing Great Britain and Russia or South Korea and Germany.

I also ask the students to find websites that list the countries in a manner that we would call productivity. This web site has a graph of the top 8 countries based on medals per athlete while this one includes the inverse of productivity in its table on athletes per medal.

The students generally find this to be a fun exercise and appreciate that inputs and outputs may not be as simple as one might think. In the course of searching for Olympic productivity on web sites, they also find sites that are interesting even though the site does not answer the productivity question. One example is this site, which explains that American workers are less productive during the Olympics. Finally, the exercise was very timely this past week since the Olympics are currently going on.

Guest Post: Using POM for Windows’ Powerful Normal Distribution Calculator

Our Guest Post today comes from Prof. Howard Weiss, at Temple University. Howard is the developer of the superb POM for Windows and Excel OM problem solving software that we provide free with our OM texts.

One of the very useful features of POM for Windows is a Normal Distribution Calculator. This tool makes these computations much easier than using a calculator. It is also more meaningful for your students than using Excel because of the graph that is presented. The Normal Distribution Calculator can be found in the Tools menu in POM or on the toolbar and is useful for solving problems in the Project Management, Forecasting, Quality Control and Inventory (safety stock) chapters.  Example 10 from Chapter 3, Project Management, is displayed below.

The students indicate the type of problem – compute a probability as in this example or find cutoff values given a probability, enter the mean and standard deviation, indicate whether the problem is one or two-tailed,  and either give the cutoff(s) or give the probability. POM displays a very readable, meaningful curve which enables the students to quickly identify the answer.