Save Dockless Shared Bikes: IOT, Big Data & Machine Learning
Dockless shared bikes are well known for their cheap prices, flexible parking, and “last-one-mile” ability. However, worldwide, a series of problems raised. Theft, destruction and casual placement are particularly prominent[1].
Out of “curiosity” , interest conflict and even antisocial psychology, there are considerable amount of people trying to steal and further destroy the bikes.
Some “normal” usages also creates troubles. As I mentioned in Signal 1, these bike are much more likely to crowd around subway stations or park entrances, which inversely worsen the public transit.
From an economic view, these problems can be concluded as lack of “supply liquidity”. When the new supply is released into a market, the market needs some time to adjust the balance of its supply and demands. Nevertheless, with IOT, Big data and Machine learning techs, we can speed up the process.
Mobike, one of the dockless bike companies, is developing models for biking usages prediction[2], and accordingly hiring workers to manually adjust bike supplies among different parking blocks. With accurate supply adjustment, less surplus bikes will be stolen, destroyed and randomly parked.
However, although the biking data are collected anonymously by IOT tech, concerns about data security and privacy are still existing[3].
References:
[1] Theft and destruction of dockless bikes a growing problem
