Third Party API References

Using Distance Matrix AI API

1. Why Distance Matrix AI API?

Initially, we considered using the Haversine formula as the main methode to calculate distances. However, the Haversine formula calculates distances along a straight line, disregarding real-world road conditions such as turns, bends, and traffic patterns. This can lead to inaccurate distance estimates.

For instance, the Haversine formula might calculate the distance between locations A and B as 700 meters. However, the actual driving distance between these locations could be 1 kilometer due to road bends.

These inaccuracies can significantly impact our project, as we limit food recommendations to a 20-kilometer radius from the user's location. Using the Haversine formula could result in recommendations beyond the intended 20-kilometer radius, potentially leading to dissatisfied users.

Therefore, we opted to utilize the Distance Matrix AI API. This API takes into account real-world road conditions, providing more accurate distance estimates and ensuring that our food recommendations remain within the specified 20-kilometer radius. This API can be said to be similar to the Google Maps Distance Matrix API (opens in a new tab). For more details, you can read this documentation about Distance Matrix AI API (opens in a new tab).

By employing the Distance Matrix AI API, we can ensure that our project delivers accurate and reliable distance information, enhancing user satisfaction and project effectiveness.

2. How We Use the API

Example using python:

def get_distance_with_distancematrix_ai(latitude1, longitude1, latitude2, longitude2):
    """Menggunakan DistanceMatrix.AI untuk menghitung jarak antara dua titik."""
    try:
        url = f"https://api.distancematrix.ai/maps/api/distancematrix/json?origins={latitude1},{longitude1}&destinations={latitude2},{longitude2}&key={DISTANCEMATRIX_API_KEY}"
        response = requests.get(url)
        data = response.json()
 
        # Log the API response for debugging
        logger.debug(f"DistanceMatrix API response: {data}")
 
        if data.get("status") != "OK":
            logger.error(
                f"Error from DistanceMatrix API: {data.get('error_message', 'Unknown error')}"
            )
            return None
 
        if "rows" not in data or not data["rows"]:
            logger.error("No rows found in DistanceMatrix API response")
            return None
 
        if "elements" not in data["rows"][0] or not data["rows"][0]["elements"]:
            logger.error("No elements found in DistanceMatrix API response rows")
            return None
 
        element = data["rows"][0]["elements"][0]
        if element.get("status") != "OK":
            logger.error(f"DistanceMatrix API element error: {element.get('status')}")
            return None
 
        distance = element["distance"]["value"] / 1000
        return distance
    except Exception as e:
        logger.error(f"Error calculating distance: {e}")
        return None

Base URL and Endpoint:

https://api.distancematrix.ai/maps/api/distancematrix/json?origins=<origin_location_1|origin_location_2|...|origin_location_n>&destinations=<destination_location_1|destination_location_2|...|destination_location_n>&key=<your_access_token>

Parameters:

• origin = user's latitude and longitude
• destinations = latitude and longitude of the recommended place
• key = your own GMAPS API KEY

Example of Response (if success):

{
  "destination_addresses": [
    "St John's Church, North End Rd, London SW6 1PB, United Kingdom"
  ],
  "origin_addresses": [
    "Westminster Abbey, 20 Deans Yd, London SW1P 3PA, United Kingdom"
  ],
  "rows": [
    {
      "elements": [
        {
          "distance": {
            "text": "7.3 km",
            "value": 7346
          },
          "duration": {
            "text": "23 mins",
            "value": 1401
          },
          "origin": "Westminster Abbey, 20 Deans Yd, Westminster, London SW1P 3PA, United Kingdom",
          "destination": "St John's Church, North End Rd, Fulham, London SW6 1PB, United Kingdom",
          "status": "OK"
        }
      ]
    }
  ],
  "status": "OK"
}

3. Advantage of Using This API

• Accuracy: Distance Matrix AI leverages real-world data to provide accurate distance calculations.
• Efficiency: The API offers a fast and reliable way to retrieve distances.
• Flexibility: The API supports various travel modes which can be incorporated in the future.

4. Disadvantage of Using This API

Due to limited development costs, we use this API with a free plan which causes the following shortcomings:

• Usage Limits: Free plans have lower usage limits in terms of the number of requests per day or per month.
• Rate Limits: The rate at which we can make requests (requests per second) might be lower we the free plan, potentially leading to slower response times for end-users if many requests are made simultaneously.

5. Alternative Solution

Because the free plan has a usage limit of 1000/month, we overcome this usage limit by creating a fallback so that if the Distance Matrix fails to execute, the distance calculation will be carried out using the logic of Haversine formula algorithm that we have created. However, as we said before, this Haversine has several drawbacks, for more details, you can read this section again.