Can Foldit change the Retail supply chain landscape?
On behalf of : Majush Koshy Philip
Will you ever approach a gamer for a planning strategy task? Or will you ask crowd to help you with retail problems?" No." if this is probably your answer, you may be wrong in near future. Recently a group of gamers playing a game called "fold it" has proved that they have the potential to replace computers and scientists by solving problems which computer took many years to solve . The question here is if Crowd (read as gamers) can replace computers in various high skilled activity, why use them for optimization of retail supply chains or planning activities in general?
Foldit is a game devised by the University of Washington Game science and Biochemistry departments. The aim of the game is to fold the protein into a chemically stable structure with lowest energy. Every kind of protein folds up into unique shape which is the most stable state it can adopt. Compare the energy of all the different ways a given protein can be folded and find the most efficient and the minimal one wins. This game is played by anyone who can register at foldit site. The best ones are shared across for further improvement. The new algorithm independently discovered by scientists and by Foldit players outperforms previously published methods. Thus, online scientific game frameworks have the potential not only to solve hard scientific problems, but also have abilities to strategize new algorithms.
What is the relevance of this in retail? Let's think in another way - What is the retailer trying to achieve? He wants to maximize or improve the profit function (revenue, profit, inventory, more delivery) minimize the cost function (Out of stock cost , inventory cost , delays , travel etc.) subject to constraints. How is a gamer trying to score points in foldit? He wants to fold the protein structures to a low energy one, adhering to the rules of chemistry. If the game can be set in such a way that it mimics the cost/profit functions without being overly explicit, it will bring out the power of crowd.
Delivery center DC1 has to deliver shipments to 30 stores in the shortest possible route; yes it is a modified travelling salesman problem. This can be modeled in different ways. Distance from each stores can be maintained as a table and Game can be thrown to crowd for optimizing the route. The art here is to hide the retail problem as a game (say ... help Tom to reach home after meeting all his friends in the shortest possible route). Only the Game designer knows Tom is the DC Truck and friends are shops it should cover.
The first advantage is that you get more diversity in ideas, opinions and solutions. This may also happen with lower cost. This is a type of crowdsourcing a.k.a crowd planning - is not a work of single person or a single thought process. The gamers involved think differently and each of them are using different algorithm to modify solution. Foldit researchers asked the winning teams - what was their logic behind acting/ folding in that particular way. Many had logical answers some had just one thing "intuition- I felt that way". This is something only human intelligence is capable of.
Are there any risks involved? Obviously no one will be willing to make the retail problem public.
This is where the Game part plays the role. This decouples the actual issue and masking the real problem with a suitable game. The 'non-public' or 'privacy' aspect depends on how well a game is devised which can prevent the world from peeping into actual issue. Remember, everybody talked about insecurity in cloud when cloud was first presented.
The time is not far when the supply chain product vendors conduct online games to improve the algorithm efficiency of their optimization techniques. Overall this got a huge scope and it all depends on how the retail community views it.