GUI for App based Passenger Pickup System
GUI for App based Passenger Pickup System
GUI for baggage screening in the vehicle
Variation of time spent at TSA using new algorithm
Motivation for our project
Neural Network implementation to predict the future behavior based on flight schedules
Model for reducing congestion and enhancing travel experience
- How frequently do we arrive early and wait long hours or arrive late to the airport?
- How often do we forget to not carry liquids above 100ml or sharp objects and had to throw stuff away at the TSA screening?
- Were we ever made to wait for more than 30 minutes for baggage check-in and security check?
What it does?
An increased number of passengers especially during the peak time period is a serious problem for airports because not all airports are designed to serve passengers beyond the threshold point which causes congestion and the number is expected to be on the rise over the coming years. Our goal is to explore ways to improve Customer Experience (CX) but still be profitable for both the airport and the airlines. Our focus areas are to enhance the travel experience from the moment the passenger booked his/her flight. This can be achieved by monitoring human behavior and communicating it in the right way to reduce congestion by using Machine Learning techniques.
How we built it?
Better communication, accommodation, and compassion will greatly improve the Customer Experience. We aim to achieve this by equipping the airline company with a fleet of semi-autonomous vehicles that have the capability to scan the baggage and weigh it while kept in the boot of the car. This way, the passenger is alerted about the presence of banned stuff like more than 100ml of liquid, sharp objects which otherwise have to be thrown away by the TSA while screening. This might be heartbreaking when passengers have to throw away valuables when unknowingly packed into the luggage. With these semi-autonomous vehicles, the presence of these substances can give the passenger an option to return it to their pickup location.
A multi-vehicle path planning algorithm schedules the order of pickup for multiple customers
A MATLAB algorithm is used to schedule the pickup time for each passenger so that not all the passengers arrive at the same time to the airport
This is achievable by a fee that a user can pay to arrive at a time closer to their flight departure time to avoid waiting at the airport for about 3 hours
The airline personnel then sends a text to the passenger and the pickup time by the airline owned vehicle
The passenger then loads their baggage into the trunk. Now, the passenger has the option to either check-in their baggage by themselves which has no fee or to opt for the enhanced passenger experience by paying a fee which speeds up their wait time at the airport cause the check-in would be done by the airline personnel as the weigh-in and baggage screening is done at the car itself
Another solution to reduce congestion that we have implemented is by using a neural network to train the algorithm using a data set that has the data for ~20 hours of passenger entry and exit for every time period throughout the day. Once the algorithm is trained, it can predict the number of passengers at each time step for the next day. Using this we can bring in the passengers at the time when there is less congestion
Challenges we ran into
- Incorporating feasible, near-future techniques as a model for our futuristic airport
- Mentors steered us in the right path to focus more on decreasing the congestion for passengers during peak hours of operation
- Few setbacks while coding
Accomplishments that we're proud of
- Machine Learning algorithm works pretty well in predicting the passenger traffic the next day using the given data set
- Modeling the enhancement of Passenger Experience by reducing wait time at the airport using tech equipped vehicles for picking up passengers
- Developed a GUI for an app-based booking system
What we learned
- There is plenty of room for development and enhancement of passenger experience
- Immediate need to reduce congestion as the number of passengers traveling at peak hours is constantly on the rise
- Price/miles just like Uber if one wants to be exactly on time the price is 2 times the normal price
- If a passenger is willing to spend time at the airport it is the normal revenue
- Carpooling option can be added for 0.5 times the normal price
- Advertising revenue can be generated through AR app or AR glasses which shall be used for navigating within the airport or to access in-flight entertainment or to customize food offered on the flight
What's next for Future of Air Travel: Enhancing the Traveling Experience
- IATA expects 7.2 billion passengers to travel by 2035
- Let’s assume our sample size is 0.1% of the total (7.2 million per annum)
- The intended initial market is individuals in the age of 18-50 in the United States who travel for leisure or business which is 1 billion as of 2018
- Based on half-a-dollar profit on each passenger, the estimated profit comes out to $7.8 million/annum
Value proposition (Total available market):
- Kate: A millennial who earns more than $100k, loves to travel and worries about the time and traffic to get to the airport. VP: She would travel using our 2nd tier plan, probably paying more
- Nate: GenX’er who’s not good with tech, wants to travel to his kid’s marriage, very careful in spending though. VP: Opt for 0.5x (probably cheaper than Uber) and opt-out of all other services we provide
- Logan: Married, expecting child, millennial and would travel for work. As the company would be paying, he’ll opt for business class if possible. VP: 2x option, arrives on time and can have the ultimate travel experience