Getting Better All the Time

By Sanjay Saigal

It's dinnertime.

You plan to grill steaks for yourself, spouse, and kid. Each steak takes 15 minutes per side. Your grill has space for two steaks at a time. How long until the three of you can sit down to dinner together?

The obvious answer is an hour. First, you grill two of the steaks. At 15 minutes per side, that takes half an hour. Then you grill the third steak. Another 30 minutes. In all, an hour.

Can you do better?

Sure you can.

Start as you did before, by grilling two steaks simultaneously. But this time, after 15 minutes, do something different: flip one steak (let's call it steak #1), set the other steak aside on a plate, and put steak #3 on the grill. After another 15 minutes, steak #1 is done, plate it. Flip steak #3 and grill that along with steak #2 (uncooked side down). In 15 minutes, all three steaks are done. Total turnaround time? 45 minutes. A 25% improvement over the obvious solution.
Now, there are no tricks involved. The first procedure, which takes an hour, has the following steps:
Steak1.side1 & Steak2.side1 
  ...cook 15 minutes...  
 Steak1.side2 & Steak2.side2 
  ...cook 15 minutes... 
 ...cook 15 minutes... 
 ...cook 15 minutes...
The better cooking procedure is:
Steak1.side1 and Steak2.side1
  ...cook 15 minutes... 
  Steak1.side2 and Steak3.side1
  ...cook 15 minutes...
  Steak2.side2 and Steak3.side2
  ...cook 15 minutes...

The second procedure is faster because it involves only three stages instead of four. Now, let's ask: Can you do better than 45 minutes?

You cannot and here's why: The meat needs a total of 90 minutes on the fire (3 steaks times 2 sides each times 15 minutes per side). At most, two steaks can be cooked simultaneously. So the best possible turnaround time is 45 minutes (90 minutes divided into 2).

The two-step described above -- finding a better way and showing that it is the best possible -- lies at the core of a discipline called Operations Research, or OR. 

The OR expert straddles business, information technology (IT), and the mathematics sciences. The business world is rich in process environments whose efficiency may be profitably improved. The OR expert transforms the "problem" into models and algorithms that can be programmed on computers to yield usable results. (In our example, we went from the world of backyard grilling to into the abstraction of sequencing "operations".) This process, as much art as science, often needs a judicious synthesis of tools from mathematics, computer science, and statistics.

Our lives are improved by OR every day. To highlight a few examples, using OR,

  • Utilities generate electricity inexpensively and deliver it reliably to your home,
  • The postal service and express shippers optimize deliveries and pick-ups on the fly, and
  • Retailers decide how much of each item to stock, how to price it, and where to display it.
Admiral Mike McMullen, America's current Chairman of the Joint Chiefs of Staff, has often mentioned the importance of his OR background. (Here Adm. McMullen talks about how OR influences his world-view. Note for hard-core Fallows fans: He even mentions China!)

Yet, mentions of OR in the popular press are few and far between. Worse yet, popular media articles covering OR manage to avoid mentioning OR entirely! This points to a severe problem for the profession, which I will take up in a later post. 

(Picture of mouthwatering steaks courtesy _bubby_@Flickr.)

Sanjay Saigal is founder and CEO of Mudrika Education, Inc., with offices in Silicon Valley, CA and Delhi, India.