IngoX

Inspiration

In Germany, delays of buses or traffic jams are very common especially during rush hour.

What it does

By controlling the traffic signals, a bus will have priority to pass the intersection. The more passengers on board, the easier to avoid traffic jams.

How I built it

Use a camera box (Raspberry pi + camera + Arduino shield) to catch traffic info at the intersection Use a paxcounter to estimate current passengers on the car Connect the hardware to TTN Get GPS info and traffic light info from the third party By using all the info above, use SUMO networks to optimize the traffic light

Challenges I ran into

Connect hardware to TTN. Make a neural network run on raspberry pi, make Sumo model run

Accomplishments that I'm proud of

Teamwork, build a car counter neural network, build Sumo simulation, build paxcounter and car counter camera box which connects to TTN

What I learned

TTN, Rabsberry, Neural Network lite, ...

What's next for IngoX

Research purpose: make data available for research

Demand Responsive Transport: passengers can request for shuttle buses from A to B. And a shuttle bus will be assigned to pick up the passenger. Different fares depend on the urgency and corresponding maximal arriving time is calculated. If a new passenger is assigned to a shuttle bus, the new route of the shuttle bus should not harm the maximal arriving time of other passengers on board.

P+R last mile service