I recall this service retirement being announced a while back and it being made clear it wasn't moving to Azure Maps. The Azure Maps team put together documentation on how to create a similar solution here: https://learn.microsoft.com/en-us/azure/azure-maps/itinerary-optimization-service
They have a live code sample here: https://samples.azuremaps.com/rest-services/mio with source code here:
- https://github.com/Azure-Samples/AzureMapsCodeSamples/tree/main/Samples/REST%20Services/MIO
- https://github.com/Azure-Samples/AzureMapsCodeSamples/tree/main/Services/SampleFunctions
Aside from that there are a lot of ways to do this in Azure already. Typically, these types of calculations are very heavy and it's best to do it as close to the data as possible (e.g. in the database). Azure Postgres has extensions such as pgRouting and PostGIS, which provide solutions for this. There are also a ton of open-source solutions for different edge case scenarios. There are also third part solutions that are cheaper:
I'm not on the maps team so can only speculate as to why this service is being retired. Nearly every customer I worked with around logistics optimizations had unique constraints/requirements that few others would be interested in. In many cases this Bing service wasn't a good fit for a solution with these unique requirements and it was better (cheaper/faster) to simply generate a network graph using a distance/route matrix API and feed that into a custom solution.