Based on Wiki definition:

“Dynamic pricing is a pricing strategy” in which businesses set highly flexible prices for products or services based on current market demands”.

Many companies, especially in domains like hospitality (i.e.hotels), traveling, and entertainment, have tried to apply this idea in order to increase their profit. However, all of them considered only the supply-demand relationship and no of them thought the use of environmental conditions. Our project exploits the idea of Internet of Things (IoT) and Intel Edison to combine supply/demand relationship with environmental conditions and to apply this idea to vending machines. Additionally, because we would like to have a win-win relationship between customers and companies, we allow users to register to our system. By using smartphones, they can find the location of the best offer, recommend offers and send gifts to friend.

Step 1: Inspiration

The following story can describe better our inspiration:

It is a very hot and dry day. Everyone sweats and begs for a cool beverage. The vending machines lurk around the city for new victims. To a remote area of a city that no stores are around, a vending machine “sees” a desperate and thirsty person walking around without being aware about the machine. Immediately, the machine tries to understand if this is a new customer by looking its memory. No, this is a new customer! The machine shout to the customer:

“Hello there!”

The customer turns around by searching the source of the voice and the machine continues:

“The weather is extremely hot today. What about a cool soda?” The customer approaches the machine slowly, pries the machine, and replies: “I really want a cool soda”.

"No problem. 3 pounds please."

"Whoa! This is too expensive!"

"But we offer a cool soda in this deserted area and with this hot weather. Otherwise you can register to our system by downloading an app to your mobile phone to get 1.5 pound off or to visit the closest and cheaper vending machine that is located in A Street 3 miles away."

The hopeless customer decide to download the Smart Vending Machine App and to register to the system and get the soda for 1.5 pounds. Otherwise, the vending machine sends the address (Google Maps location) of the closest and cheaper vending machine.

We also assume that our smart vending machines can sell alcohol only to the registered users that we know their ages. When the register customer approaches a machine and intents to buy alcohol, the machine ask from him/her to blow toward the machine. If the registered customer is sober, the machine gives the beer to the customer. Otherwise the machine barks to make the customer to dissuade from buying alcohol. If the customer insists and start tilting the machine, it calls the police.

After some days, the customer decides to explore the features of the Smart Vending Machine App. The first impression was that the application provides only prices of the closest vending machines. However, the customer realizes that he/she can set alarms of the soda prices, send recommendation of a vending machine, or a gift to his/her friends around the world through Viber, Skype, WhatsApp, Facebook, etc.

To sum up, our smart vending machine system:

Attracts new customers by exploiting motion detection and producing audio and visual messages.
Keeps the customers by personalizing vending machines (personalized sound messages, price alerts and notification).
Intrigue more new customers by using the already registered customers who send recommendations and gifts.
Makes profit by applying dynamic pricing distributed algorithm (local execution in vending machine) based on Supply Demand, and Environmental temperature.
Defines customers behavior by pushing the vending machine data to Intel Cloud.
Keeps our registered users sober!!!

Overall, our system ensures the win-win relationship between customers and companies.

Step 2: Design

Within the time constraints of Intel Hackathon we achieved to develop (see above Figure):

The Brain of the vending machine based on Intel Edison which executes the dynamic pricing algorithm locally and incorporate all the different sensors (except from microphone - we assume that the user give commands from the keypad) and output devices.
Dynamic pricing algorithm based on demand (rate of buying), supply (current stock), and temperature.
The smartphone App with which the registered customers can send recommendation and gifts to others by exploiting Skype, Viber, QR codes, WhatsApp, SMS, etc.
Communication between smart vending machine and smartphones through Bluetooth and WiFi.
Communication between smart vending machine and Intel Cloud to verify the results of our algorithm.