OpenRouteService Matrix API: ‘6004 Request Parameters Exceed Server Limits’ Even When Under 3,500 Routes

SDKs R
1

I am using OpenRouteService (ORS) Matrix API in R to compute travel times between grid centroids (origins) and hospitals (destinations). However, I keep encountering the error:

Error: Openrouteservice API request failed [6004] Request parameters exceed the server configuration limits. Only a total of 3500 routes are allowed.

Despite ensuring that my total routes (origins * destinations) do not exceed 3,500, I still receive this error.

My Setup

I have: • 40 origins (grid centroids) • 85 destinations (hospitals) • Total routes: 40 * 85 = 3400 (should be under 3,500).

I have a total of 585 origins, however, I’m batching my request to use 40 origins at a time (see batch_size in the code below).

Please see my code below.

# Ensure `grid_coords` is a matrix for `nrow()`
if (!is.matrix(grid_coords) && !is.data.frame(grid_coords)) {
  grid_coords <- do.call(rbind, grid_coords)  # Convert list to matrix
}

if (!is.matrix(hospital_coords) && !is.data.frame(hospital_coords)) {
  hospital_coords <- do.call(rbind, hospital_coords)  # Convert list to matrix
}

# ORS API limits (adjust if needed)
batch_size <- 40  # Max centroids per batch

# Initialize results dataframe
results <- data.frame(
  grid_id = integer(),
  nearest_hospital = character(),
  nearest_hospital_lon = numeric(),
  nearest_hospital_lat = numeric(),
  travel_time_min = numeric(),
  stringsAsFactors = FALSE
)

for (i in seq(1, nrow(grid_coords), by = batch_size)) {
  batch_end <- min(i + batch_size - 1, nrow(grid_coords))
  centroid_batch <- grid_coords[i:batch_end, , drop = FALSE]  # Matrix format

  num_origins <- nrow(centroid_batch)
  num_destinations <- nrow(hospital_coords)
  total_routes <- num_origins * num_destinations

  print(paste("Batch", i, ": Origins:", num_origins, "| Destinations:", num_destinations, "| Total Routes:", total_routes))

  # Convert `centroid_batch` and `hospital_coords` to list format for ORS
  centroid_list <- split(centroid_batch, seq(nrow(centroid_batch))) %>% 
    lapply(function(row) c(row[1], row[2]))

  hospital_list <- split(hospital_coords, seq(nrow(hospital_coords))) %>% 
    lapply(function(row) c(row[1], row[2]))

  # Combine origins and destinations into ORS-compatible format
  locations <- c(centroid_list, hospital_list)

  # Send batch request to ORS Matrix API
  travel_matrix <- ors_matrix(
    locations = locations,
    profile = "foot-walking",
    metrics = c("duration"),
    units = "m"
  )

  # Extract relevant travel time matrix
  travel_time_matrix <- travel_matrix$durations[1:num_origins, (num_origins + 1):(num_origins + num_destinations)] / 60

  # Find nearest hospital by shortest travel time
  nearest_hospital_index <- apply(travel_time_matrix, 1, function(x) ifelse(all(is.na(x)), NA, which.min(x)))

  # Store results
  batch_results <- data.frame(
    grid_id = i:batch_end,
    nearest_hospital = ifelse(is.na(hospitals$name[nearest_hospital_index]), "Unknown", hospitals$name[nearest_hospital_index]),
    nearest_hospital_lon = hospital_coords[nearest_hospital_index, 1],
    nearest_hospital_lat = hospital_coords[nearest_hospital_index, 2],
    travel_time_min = apply(travel_time_matrix, 1, min),
    stringsAsFactors = FALSE
  )

  # Ensure `batch_results` has the same columns as `results`
  if (identical(colnames(results), colnames(batch_results))) {
    results <- rbind(results, batch_results)
  } else {
    warning("Column mismatch detected, skipping batch.")
  }
}

# Merge results with grid centroids
grid_centroids <- grid_centroids %>%
  left_join(results, by = c("row.names" = "grid_id"))
2 Likes
2

Thanks for reaching out and for providing a detailed description of your issue :heart:

The problem you are encountering is due to the missing sources and destinations parameters in the call to ors_matrix. Without these a full 125x125 matrix between all of the provided locations is computed. You might want to modify your script to have sth along the following lines.

Cheers,
Andrzej

# Send batch request to ORS Matrix API
  travel_matrix <- ors_matrix(
    locations = locations,
    sources = 1:num_origins,
    destinations = (1:num_destinations) + num_origins
    profile = "foot-walking",
    metrics = c("duration"),
    units = "m"
  )
2 Likes
3

Thank you, Andrzej, I really appreciate your help. I’ve updated the code with sources and destinations. However, I’m getting the following error:
"Parameter ‘destinations’ has incorrect value or format.

4

You’re very much welcome! :heart:

Thanks for your feedback. My suggested solution might actually have a bug, sorry for that :innocent:
I believe the indices provided in the API call should be 0-based (unlike in R), so please try shifting them by 1 as in the following updated code snippet. I hope this solves the issue!

Cheers,
Andrzej

# Send batch request to ORS Matrix API
  travel_matrix <- ors_matrix(
    locations = locations,
    sources = 1:num_origins - 1,
    destinations = (1:num_destinations) + num_origins - 1
    profile = "foot-walking",
    metrics = c("duration"),
    units = "m"
  )
2 Likes
5

Thank you very much, Adrzej! This solved the issue. Really appreciate your help and timely response! Love the community aspect of this!

2 Likes
6

I just cited you (your ORS matrix documentation) in my master’s thesis. I’ll have to send you a copy when I’m done. My thesis is on healthcare facilities accessibility for prenatal care in N’Djamena, Chad.

4 Likes
7

@ka88
Congrats :tada:
Looking forward to checking out your thesis!

2 Likes