jguthula's Diary

Recent diary entries

English version:

Amazon has been an active contributor to OpenStreetMap (OSM) in Germany for over a year now. The drivers of our delivery service providers at times encounter problems where OSM data is not matching with on-ground reality (e.g., where the map suggests a left turn, but the left turn is legally restricted as per the street signs). These discrepancies are reported by the drivers, but due to absence of ground truth (street images), we are unable to investigate and rectify the map defect in OSM. These defects risk the safety of not just drivers of our delivery service providers, but all users who might be using OSM as a base map for navigation purposes. In order to overcome the lack of ground truth in Germany, Amazon has signed up for a short term contract with CycloMedia Technology B.V., a private street imagery provider. Details about CycloMedia can be found here.

How are we going to utilize the data?

Amazon will investigate the locations of feedback shared by drivers spread across Germany using CycloMedia street images on CycloMedia’s web-access portal. The images will be mainly used to add restrictions related to turns, access, directions, weight, and height to improve OSM map data. When making these edits on OSM, we will attribute the anonymized source of the image along with its date of capture. Additionally, staying true to maintain the quality of data in OSM, Amazon will contribute edits to OSM directly only when we have a street image of 2019 (or more recent). However, when we find images of 2018 or older (which reinforce the feedback from the driver), Amazon will share notes or create map roulette tasks. In these situations, we request support from the German OSM community to confirm the ground reality before making an edit. Furthermore, any comments on these changesets from the community members will be taken up on priority and responded to at the earliest.

What images are we using?

We will be using parallax free 360° panoramic photographs (Cycloramas) of street images captured by CycloMedia at every 5m intervals. These high resolution spherical panoramas are located to an accuracy within 10 cm. How will we attribute the source for the edits made on OSM? All the edits on OSM made using the street images of CycloMedia will have the anonymized source ( eg.-“source = private supplier”) and the year of image captured mentioned ( eg.-“Year = 2019”) against them.

Can we share CycloMedia images with community?

No, Amazon is not allowed to make CycloMedia image material available to public or to third parties. However, Amazon and its affiliated companies are entitled to use CycloMedia Image material to make temporary or permanent edits on OSM which may be made permanently available to the public without any reference to Supplier.

German Version:

Amazon ist seit über einem Jahr ein aktiver Mitwirkender bei OpenStreetMap (OSM) in Deutschland. Die Fahrer unserer Zustelldienste stoßen manchmal auf Probleme, wenn die OSM-Daten nicht mit der Realität vor Ort übereinstimmen (z.B. wenn die Karte eine Linkskurve vorschlägt, die Linkskurve aber gemäß den Straßenschildern gesetzlich eingeschränkt ist). Diese Diskrepanzen werden von den Fahrern gemeldet, aber aufgrund des Fehlens der Bodenwahrheit (Straßenbilder) können wir den Kartenfehler in OSM nicht untersuchen und beheben. Diese Defekte gefährden nicht nur die Sicherheit der Fahrer unserer Zustelldienste, sondern aller Benutzer, die OSM als Basiskarte für die Navigation verwenden könnten. Um den Mangel an Grundwahrheit in Deutschland zu überwinden, hat Amazon einen kurzfristigen Vertrag mit CycloMedia Technology B.V., einem privaten Anbieter von Straßenbildern, abgeschlossen. Details zu CycloMedia finden Sie hier.

Wie werden wir die Daten nutzen?

Amazon wird die Standorte der Rückmeldungen untersuchen, die von den über Deutschland verteilten Fahrern mit Hilfe von CycloMedia-Straßenbildern auf dem Web-Zugangsportal von CycloMedia ausgetauscht werden. Die Bilder werden hauptsächlich verwendet, um Einschränkungen in Bezug auf Kurven, Zugang, Richtungen, Gewicht und Höhe hinzuzufügen, um die OSM-Kartendaten zu verbessern. Wenn wir diese Bearbeitungen auf OSM vornehmen, werden wir die anonymisierte Quelle des Bildes zusammen mit dem Aufnahmedatum angeben. Um die Qualität der Daten in OSM aufrechtzuerhalten, wird Amazon nur dann direkt Änderungen an OSM vornehmen, wenn wir ein Straßenbild von 2019 (oder neuer) haben. Wenn wir jedoch Bilder von 2018 oder älter finden (die das Feedback des Fahrers verstärken), wird Amazon Notizen austauschen oder Kartenroulette-Aufgaben erstellen. In diesen Situationen bitten wir die deutsche OSM-Community um Unterstützung, um die Realität vor Ort zu bestätigen, bevor wir eine Bearbeitung vornehmen. Darüber hinaus werden alle Kommentare der Community-Mitglieder zu diesen Änderungssätzen mit Priorität aufgenommen und frühestens beantwortet.

Welche Bilder verwenden wir?

Wir werden parallaxenfreie 360°-Panoramafotos (Cycloramas) von Straßenbildern verwenden, die von CycloMedia in Abständen von 5 m aufgenommen werden. Diese hochauflösenden sphärischen Panoramabilder werden mit einer Genauigkeit von 10 cm aufgenommen. Wie werden wir die Quelle für die auf OSM gemachten Bearbeitungen zuordnen? Alle Bearbeitungen auf OSM, die mit den Straßenbildern von CycloMedia gemacht wurden, werden mit der anonymisierte Quelle ( z.B. “Quelle = privater Lieferant”) und dem Jahr des aufgenommenen Bildes ( z.B. “Jahr = 2019”) versehen.

Können wir CycloMedia-Bilder mit der Gemeinschaft teilen?

Nein, Amazon ist es nicht erlaubt, CycloMedia-Bildmaterial der Öffentlichkeit oder Dritten zur Verfügung zu stellen. Allerdings sind Amazon und seine verbundenen Unternehmen berechtigt, das CycloMedia-Bildmaterial für vorübergehende oder dauerhafte Bearbeitungen auf OSM zu verwenden, die der Öffentlichkeit ohne jeglichen Hinweis auf den Lieferanten dauerhaft zur Verfügung gestellt werden können.

Regards, Jothirnadh

Help required for adding access information to gates/ gated communities

Posted by jguthula on 12 March 2020 in English (English). Last updated on 14 March 2020.

Dear OSM Community,

At Amazon, we have observed that ~94% of gates mapped in OSM do not have access information updated (based on study in Florida). This poor coverage of gates lead to routing issues such as allowing general public to restricted areas; specially to gated communities (find the below example for reference). We plan to take up the task of updating the access information for gates in the US. Picture 1 OSM link:

In the above example highlighted polygon contains a gated community with 3 gates situated at three different places. Red points indicate restricted access (residents only) and green point indicates access allowed for general public/ deliveries along with the local residents.

How will we get the access information about gates and add it to OSM?

Whenever our delivery partners route through publicly non-accessible gates/ gated communities, they will collect the access information for those gates and share it with mapping team. Our mapping team will only add access information to gates if both - drivers feedback and OSM data sources match same information. If there is no supportive data, our editing team will drop a note for local community to add appropriate access information using local knowledge. How to add access information to gates in OSM? In our research, we observed most of the access information added to gates are using the “key: access”. Though there is not much coverage of access information to explore for gates, the different gate access scenarios we identified that can be added are:

Key Value Description
access no Gates with permanent closure using boards of “do-not-enter”
access private Gates with which general public or deliveries are not allowed through
access permissive Gates open to general traffic until the time owner revokes the permission
access delivery Gates from which only ‘deliveries’ are allowed
access discouraged A legal right of way exists but usage is officially discouraged. Will only add this information if a traffic sign exists

Please do let us know if there are alternative ways of adding access information to OSM dataset. If you are aligned to our approach, we would like to initiate adding ‘gate access’ in Florida followed by other parts of the US.

Dear OSM Community,

At Amazon we are coming across a situation where unmaintained track roads have no access tags. We would like to know your suggestions/feedback on how to tag these roads.

Track Roads:

By definition track roads implies traversal of motor vehicles and are mostly used for agricultural use. We have private GPS traces showing where delivery vehicles go, and when we match them against OSM data, we see they’re going along highway=track ways. We want to add appropriate access tags so that routers can route correctly along the data in the future.

Screen Shot 2019-07-16 at 9 57 15 PM

The driver trace is on-the-ground evidence that the road can be used by motor vehicles for deliveries, but we have no evidence either way that the road can be used for general traffic.

There are four different ways we’ve come up to tag this information:

1. motor_vehicle = yes

Our editing team used to do this, but the community has pointed out that it’s incorrect because we don’t know if the roads can be used for public motor vehicle traffic. This is particularly a problem for track roads that could connect two roads.


If the track road were tagged as motor_vehicle=yes, the data would show that anyone could connect from one side to the other. We don’t have evidence on that.

2. motor_vehicle = destination

This indicates that local traffic can go there. Unfortunately, it also says through traffic can’t use it, which we also have no evidence on apart from our delivery partners route information.

3. motor_vehicle = delivery

This is the same as #2, except it indicates delivery only. All public traffic will be blocked from routing via this road.

4. service = driveway + surface = unpaved

Since most of the track roads are leading to independent houses or commercial entities, these track roads can be converted to driveways and add ‘surface=unpaved’ information as per available sources. This might not be applicable to all the track roads but only to roads leading to housing entities and vehicles visible in satellite images.

Our personal preference is the combination of #2 and #4 as we think this correctly captures the information we have.

What do community think is the most appropriate tag for this case?

As per community request, we’ve stopped adding motor_vehicle=yes to track roads and reverted all the changes we made previously in UK while we sort this out. Do let us know your suggestions.