Bhuvan is the new Earth

Bhuvan is ISRO's (Indian Space Research Organization) answer to Google Earth. Although why they needed to answer Google is beyond me. Anyway the news so far on Bhuvan is very good. It can zoom to levels of up to 10 mt compared to Google Earth's 200 mt. It provides multilayer information, images are upgraded every year (compare with four years for Google Earth) and has an option of viewing on different dates.
The link to download the software is http://bhuvan.nrsc.gov.in/; but like all things Indian the website is currently not responding due to excessive traffic. They can launch satellites and develop sophisticated software, but some how they can't handle traffic on their site. Tell me if you can get through and download the software.

Yebol: Seems like a good idea.

Yebol; the new search engine is fast (obviously), is more robust than Bing in my opinion (i couldn't find relevant results to certain terms....for example try searching for "onesmartidea blog")and certainly smarter than Google (because of a higher results to noise ratio)but not as robust. Yebol promises to get better though in the next 3 to 6 months. It seems to me to be based on a Neural Networks approach to search.
If you ask me (although I wonder why?) the jury is still out as far as the best search experience is concerned, but Yebol is definitely something to look out for.

Yebol: A new search engine

A new semantic search engine called Yebol has recently been launched. It claims to be able to provide categorized results for over 10 million search terms.
You guys might wanna go check it out and tell me about your experience.

The Pursuit of Happyness

I had this discussion with a friend of mine a couple of months ago about happiness (I hope she doesn't mind me using our conversation as a piece here). Now; for the sake of context she is currently a freelance writer and is doing well but not astoundingly well yet. Before becoming a freelancer, she was an editor with one of India's biggest (or best or wotever...) KPOs. Although being a freelance writer is what she always wanted to be, she was bothered by the thought that in 5 years, she might realize that her career is going nowhere and would end up at some KPO or some other 9-5 job and would be unhappy and bitter considering the fact that she would have been in a higher position at a similar place if she hadn’t quit her job. Her situation is similar to the dilemma faced by many people at some point of their careers. Here’s what I think.

For the sake of simplicity assuming my friend Brinjal (No that’s not her real name) puts her honest efforts towards her attempts at being a freelancer (counting each project as an attempt) the probability of her being successful at an attempt is Pa=0.5. The probability calculation is highly simplified and does not consider micro factors. This is considering that an attempt shall either be successful or unsuccessful and then probability is calculated as one calculates the probability of either heads or tails as an outcome of a coin flip. Similarly the probability of her being unhappy or bitter if at all she has to go back to a 9-5 job is also Pb=0.5. This probability is calculated considering that she can be either happy or unhappy. This probability calculation is actually more accurate than we think. When we look at the situation in light of the synthetic happiness concept, we shall realize this. For details on the same check out this video on YouTube Dan Gilbert: Why are we happy? Why aren't we happy? (the video is in fact of far greater consequence than anything I am saying here).

The combined probability of Brinjal being unhappy = Pa x Pb = 0.5 x 0.5= 0.25.

This means that there is a 25% chance of her being unhappy. Now consider a scenario where she makes multiple attempts. The probability of success in each case being Pi=0.5 (i= 1,2,3…….n). Then the combined probability of Brinjal being unhappy= Pa x Pi x Pb (i=1,2,3……n). Consider a case where she makes only three more attempts after the initial attempt. Probability of unhappiness = Pa x P1 x P2 x P3 x Pb= 0.5 x 0.5 x 0.5 x 0.5 x 0.5= 0.03125= 3.125%. More attempts she makes less are her chances of being unhappy. Intuitively as well we understand that Brinjal is better off making multiple attempts to fulfill her dream if she doesn’t achieve at her first attempt what she set out to.

What has all this got to do with Supply Chain Management? Well….nothing really. I shall get back to SCM the next time.

Surely a very smart idea

Can't really explain this one. You have to check it out http://www09.wolframalpha.com/. Seems to me to be one of the best things to have happened on world wide web in recent times. Check out the introductory video and you shall go crazy. I can't begin to imagine the possibilities.

You also might want to check out Microsoft's latest endeavor in search http://www.bing.com/.

Revenue Management

My theme for the week is revenue management. Revenue management, as a lot of you know is the practice of differential pricing for different market segments with the aim of maximizing profits. A common example is pricing of airline tickets. Most airlines have a pricing structure; wherein they charge a lesser amount to people who are willing to book their tickets in advance and higher price is charged to people who need to book at the last minute. It is very interesting how an airline company in such a scenario allocates its capacity (number of seats) to the two different segments, especially when we look at the scenario in light of demand uncertainty. If the airline knows exactly how many customers would need to book their tickets at the last minute and would be willing to pay a higher price for the same, it’s obvious that the airline would want to reserve that many seats for the higher price segment and the rest of the seats it could sell at a lower price which other customers are willing to pay. This kind of an arrangement makes sense as it would maximize revenue as well as profits because the costs for an airline are mostly fixed costs and the variable costs are a small fraction of the total costs (like the cola and the peanuts they may or may not serve you during the flight).

Although an ideal situation, it is not practical by any stretch of imagination. Usually demand is uncertain and the demand from the high price segment arrives later in time in comparison to demand from the low price segment. How does an organization then deal with such uncertainty in demand? To answer the question we consider a model for allocating capacity to a segment under uncertainty (Chopra, et al., 2006). Consider only two segments (for the sake of simplicity) and the price paid by the lower price segment to be “p1” and the price paid by higher price segment to be “p2”. Also assume that the demand from the higher priced segment is normally distributed and has a mean “M” and a standard deviation “σ”. Now if we allocate a capacity of “Ch” for the high price segment the total revenue expected from the high price segment (R) would be:

Probability (demand from high price segment > Ch) x p2.

The capacity Ch should be allocated such that the expected revenue from the high price segment (which would materialize later in time) is equal to the current marginal revenue from the lower price segment, i.e. (R=p1).

In our case we would hence have the equation:

Probability (demand from high price segment > Ch) x p2 = p1.

Thus,

Probability (demand from high price segment ≤ Ch) = 1- p1/p2.

From this equation one can easily calculate Ch considering the assumption of demand from the higher price segment to be normally distributed. Using MS-Excel the calculation can be done as follows:

Ch = Norminv (1-p1/p2, M, σ).

A similar approach can be taken even if there are more than two market segments and we know the demand for each of those segments to be normally distributed. We would only need to obtain a set of reservation which would be nested. In reality however, there are multiple price slabs and demand is sometimes not normally distributed. Also the demands from different segments may not be independent and there are seasonal variations to consider as well. There are specialized software that are employed by airline companies to forecast demand and to solve the differential pricing problem.

How is all this of any use to you? (I am assuming you don’t own an airline company). You may be a freelancer who has a few regular clients who you charge a lesser price (p1) and stream of irregular clients who are willing to pay you a higher price (p2) and you can use this philosophy to maximize your revenue. The capacity in this case could be your time or any other resource (that is limited) that you use. You may be a small business owner and you could use differential pricing. And finally you may be a student or any average person and might find this interesting from a purely academic perspective.

Reference:Chopra S., Meindl P., Supply Chain Management: Strategy, Planning, and Operation (3rd edition), Pearson/Prentice Hall, NJ, 2006.

Square Zero

I have been thinking about this for a while……for about 42 seconds now. I think I want to start this blog with a theme based on Supply Chain Management, or to be more accurate some of my thoughts on the subject. I know I am no expert, but this is only a way of sharing my thoughts with you (I am hoping someone would read this…… Now that’s being presumptuous, but it’s good to be a little optimistic whenever you want to start something that’s not well thought out).

Oh! By the way…..don’t expect me to stick to SCM (or any other topic for that matter). I shall wander from topic to topic without any notice or provocation from time to time as my thoughts wander. You have to keep in mind that this is not a serious dissertation on the subject; it’s just a random blog.

“All those who wander are not lost.” - J.R.R Tolkien