System requirements

Functional:

1. List the types of vehicles the parking lot can accommodate(dimension specification)
2. Available lot for each vehicle type
3. Price per hour for a vehicle type
4. Occupied and Available space count
5. Vehicle identity
6. Vehicle to space mapping

Non-Functional:

1. Safety
2. Efficiency
3. Usability

Capacity estimation

Assumption, total available space = 1000

3 types of parking space 1. Car 2. Truck 3. TwoWheeler

Car space = 500

Truck space = 200

TwoWheeler = 300

Daily Active Users count - 100K

Request Per Second = 1

Read = Write

Storage capacity = 100 bytes of info per space

1000*100 = 1 MB storage

API design

• GET /show_map - returns the map with following information

[{

space1:{status:"Free", park_type:"Car"},

space2:{status:"Occupied", park_type:"Truck"

}]

• GET /recommendation/{:vehicle_type}

returns the recommendation list

[{

rank:1, floor:1, park_no:16

},

{

rank:2, floor:2, park_no:22

}

]

• POST /place_order

{park_no:22}

• POST /confirmation

{

park_no:22

vehicle_id:2000

}

• POST /complete_order

{

vehicle_id:2000

}

Database design

Choice of DB - SQL

inventory

• park_no:int(PK)
• space_specification:
• status:Free/Occupied
• CostPerHour
• star rating

parking

• user_id:
• park_no
• order_id
• bill_details
• vehicle_details

High-level design

Assumption:

System has 3 types of parking space

1. Car

2. Truck

3. TwoWheeler

Core Entities:

1. Authentication Service
2. Inventory Service
3. Parking Service
4. Recommendation Service
5. Billing service
6. Payment Service

Authentication Service:

1. User can create an account and use the application or use guest login facility.
2. If created account, login API validates the user
3. Manages profile information for the user.

Inventory Service:

1. Display the area map with information like park space type, occupied or free with cost information
2. Overall availability count.
3. Availability count for each park type

Recommadation Service:

Receives vehicle type criteria from user, this service talks to inventory service to match vehicle type with park type and based on availability runs a ranking algorithm and provides a recommendation to User.

Parking Service:

User chooses a park space and makes book request, this service processes the request and assigns the park space chosen by user.

Use can also use random pick, in that case parking service, talks to recommendation service and gets the rank 1 data and assign it to user.

This service also ensures the vehicles are correctly parked in the chosen park space, with a confirmation request from park space.

Also handles the exit request of user and makes the park type status back to available.

Billing Service:

Upon exit request, calculates the bill amount and send notification to user, Pay option is enabled in APP.

Payment Service:

Handles payment request from user with integration of Payment service provider.

Request flows

1. User: Able to see the whole parking space map.
2. User: specify criteria and check for recommendation
3. Place order by choosing a park space from recommendation
4. User has random pick option for the system to decide the space and return the park number to user.

Detailed component design

Parking service plays a crucial role in deciding which parking space to chose, by talking to recommendation service, by running a matching algorithm.

This algorithm - considers features like vehicle dimension, nearby space vehicle dimension, nearby slot occupied or free information, taking into account all this information

If he is a regular customer, algorithm also considers previous allotted space and try to provide the same based on availability

Trade offs/Tech choices

1. Can use redis cache to store availability information for low latency

Failure scenarios/bottlenecks

1. handle down service failure with circuit breaker pattern and appropriate fallbacl

Future improvements

1. improved recommendation algorithm
2. notification service
3. if no space available, suggesting, nearest available park service.