Directions

The HERE SDK provides a full-fledged RoutingEngine to calculate the best route directions from A to B, including multiple waypoints and localizable maneuver instructions for each turn.

Specify your preferences by setting the desired route type (fastest or shortest) and various route options (such as speed profiles, route restrictions, vignette options, and more) to find the perfect route that saves the most energy: With our advanced routing technology and our dedicated EV route planning support HERE helps you to make the planet cleaner and safer.

Feature overview:

• Calculate routes: Calculate routes with multiple waypoints for various transport modes.
• Isoline routing: Calculate isoline polygons to represent the area of reach from a given point based on time, distance or fuel consumption.
• Search along a route: Search for places along an entire route (this feature is described in the Search section).
• Import routes / route matching: You can import routes from other APIs.
• Offline routing: When no internet connection is available, you can switch to a dedicated offline routing engine to calculate routes on already cached map data or preloaded offline maps data.

Get Routes

The HERE SDK supports the following route types:

• Car route directions.
• Taxi route directions.
• Pedestrian route directions.
• Bicycle route directions.
• Truck route directions with highly customizable truck options.
• Scooter route directions.
• Bus route directions.
• Routes for electric vehicles to find the nearest charging stations (based on the calculated energy consumption and battery specifications).
• Public transit routes with highly customizable transit options (via TransitRoutingEngine).

Each route type is determined by one of the available route options such as CarOptions or TruckOptions: Find here all available route options for each supported transport mode. These options can be set to the available overloads of the route engine's calculateRoute() method.

Start your trip by creating the route engine in this manner:

try {
    routingEngine = new RoutingEngine();
} catch (InstantiationErrorException e) {
    throw new RuntimeException("Initialization of RoutingEngine failed: " + e.error.name());
}

Creating a new RoutingEngine instance can throw an error that we have to handle as shown above. For example, such an error can happen when the HERE SDK initialization failed beforehand.

As a next step, you can calculate the route based on two waypoints - a starting location and a destination (both of type Waypoint that holds a GeoCoordinates instance). Below we set default CarOptions to calculate a route that is optimized for cars:

Waypoint startWaypoint = new Waypoint(startGeoCoordinates);
Waypoint destinationWaypoint = new Waypoint(destinationGeoCoordinates);

List<Waypoint> waypoints =
        new ArrayList<>(Arrays.asList(startWaypoint, destinationWaypoint));

routingEngine.calculateRoute(
        waypoints,
        new CarOptions(),
        new CalculateRouteCallback() {
            @Override
            public void onRouteCalculated(@Nullable RoutingError routingError, @Nullable List<Route> routes) {
                if (routingError == null) {
                    Route route = routes.get(0);
                    showRouteDetails(route);
                    showRouteOnMap(route);
                    logRouteViolations(route);
                } else {
                    showDialog("Error while calculating a route:", routingError.toString());
                }
            }
        });

You can call calculateRoute() multiple times. For example, you can call it to calculate routes with different routing options in parallel.

Each route calculation will be performed asynchronously. You will get a Route list or a RoutingError that holds a possible error when completed. If all goes well, RoutingError is null. In case of an error, the route list is null. For example, the engine cannot calculate routes if a route is not feasible for the specified mode of transportation.

The showRouteDetails()-method from the code snippet above is used to show more route details including maneuver instructions. You can find the full source code in the accompanying example app. Maneuver instructions are also explained in greater detail below. The showRouteOnMap()-method contains an example, how to render a route on the map. We will explain this shortly in the section below.

You can detect possible route notices with the following method:

// A route may contain several warnings, for example, when a certain route option could not be fulfilled.
// An implementation may decide to reject a route if one or more violations are detected.
private void logRouteViolations(Route route) {
    for (Section section : route.getSections()) {
        for (SectionNotice notice : section.getSectionNotices()) {
            Log.e(TAG, "This route contains the following warning: " + notice.code.toString());
        }
    }
}

Use Route Options for Each Transport Mode

For the example above, we have set a new CarOptions instance to calculate a car route. You can calculate routes for other transport modes by using dedicated route options for each transport mode. The following route options exist:

• CarOptions to calculate car routes: route.requestedTransportMode is TransportMode.car.
• TruckOptions to calculate truck routes: route.requestedTransportMode is TransportMode.truck.
• PedestrianOptions to calculate routes for pedestrians: route.requestedTransportMode is TransportMode.pedestrian.
• EVCarOptions and EVTruckOptions to calculate routes for electric vehicles: route.requestedTransportMode is TransportMode.car or TransportMode.truck.
• ScooterOptions to calculate routes for scooters: route.requestedTransportMode is TransportMode.scooter.
• BicycleOptions to calculate routes for bicycles: route.requestedTransportMode is TransportMode.bicycle.
• TaxiOptions to calculate routes for taxis: route.requestedTransportMode is TransportMode.taxi.
• BusOptions to calculate routes for buses: route.requestedTransportMode is TransportMode.bus.
• TransitRouteOptions to calculate routes for public mass transit (only available via TransitRoutingEngine): route.requestedTransportMode is TransportMode.publicTransit.

The TransportMode is set by the routing engine after route calculation has been completed - while getRequestedTransportMode() determines the transport mode from the route options, each Section of a Route can have a different transport mode: route.getSectionTransportMode() provides the actual SectionTransportMode that has to be used for a particular section - in addition to the transport modes from above, it lists also transport modes such as ferry and carShuttleTrain.

By default, when passing only start and destination waypoints, the resulting route will contain only one route Section. Each route object can contain more route sections depending on the number of set waypoints and transport modes. Sections act as a route leg that break a route into several logical parts. Find more about this here.

Although the main transport mode is specified by the user before calculating the route, the final transport modes are set by the routing engine per Section.

All of the available route options allow you to further specify several parameters to optimize the route calculation to your needs.

Each of the above options contains a field that holds a common RouteOptions object. This option allows you to specify common options such as the number of route alternatives or the OptimizationMode to find the optimal route based on travel time and route length.

Alternatively, you can change the algorithm. For example, if you want to reach your destination quickly and the length of the route is less important to you, select the FASTEST route mode via RouteOptions. Select SHORTEST if you prefer a shorter route, and time is not so important.

To find the best route for you, the routing algorithm takes into account many different parameters. This does not mean the algorithm will always provide the absolute shortest or fastest route. For example, consider the following road network:

When you plan a trip from A to B, you may have the choice between four different roads. Let's assume that the green route represents a highway, then this road may be the fastest route, although it is longer than any of the other routes that would guide you through a city.

If you prefer to take the shortest route, the algorithm may favor the blue route, although the yellow and the red routes are shorter. Why is this so? The yellow road is the shortest route, of course, but it has to cross a river where a ferry has to be taken. This could be regarded by the algorithm as time-costly. As a result, it is possible to prefer the red or blue route instead of the yellow route, even though both are slightly longer.

Let's explore the other two options. When comparing the blue and the red route, the routing algorithm may recommend the blue route as the shortest although it is slightly longer than the red route: several turns are typically not beneficial to a driver. In this case, the red route contains more turns than the blue route, but the blue route may be the preferred route as it is only slightly longer. The routing algorithm penalizes turns and many other road properties, such as traffic lights or rail crossings that can slow a driver down.

While the resulting routes are optimized based on certain criteria, there may be situations where you don't want to rely on that. Imagine a city trip through Berlin - finding the fastest or shortest route may not be an option if you want to experience many sightseeing spots the city has to offer. In such cases, setting additional waypoints may be a good idea. Find an example below.

Get Truck Routes

The HERE SDK supports different transport modes (see above) for route calculation. Similar to the example above that shows how to calculate a route optimized for cars, you can also calculate routes for other transport types such as trucks.

While you can already get a route optimized for trucks by setting the default options, there are many more options available to find the best routes for a truck.

For example, TruckOptions contains additional fields such as TruckSpecifications to specify the dimensions of your truck, and more, that can be optionally applied to take the vehicle weight and other parameters into account.

More on truck routing can be found in the Truck Guidance section.

Get Public Transit Routes

Use the TransitRoutingEngine to calculate public transit routes from A to B with a number of waypoints in between. You can find a PublicTransit example app on GitHub that shows how to do this.

Shape the Route with Additional Waypoints

By default, when setting only start and destination waypoints, the resulting route will contain only one route Section. Each route object can contain more route sections depending on the number of set waypoints. Sections act as a route leg that break a route into several logical parts.

Waypoints are coordinates that can be set by the user to determine more sections or the shape the route (which can be useful if you want to make sure you cross a river with a ferry, for example).

Note that the online RoutingEngine supports only a limited number of waypoints of around 200 waypoints at most.

There can be two types of waypoints:

• STOPOVER: The default waypoint type. It is guaranteed that this point will be passed, therefore it appears in the list of maneuver instructions, and splits the route into separate route sections.
• PASS_THROUGH: May not appear in the maneuver instructions list and is rather treated as a hint to shape the route, for example, as a result of a touch input. This type will not split the route into separate sections.

When creating a new Waypoint object, the STOPOVER type is set by default - and this must be the type used for the first and the last waypoint. With just two waypoints acting as the start and destination, a route might look like below:

The RoutingEngine can handle multiple waypoints. The underlying algorithm will try to find the best path to connect all waypoints while respecting the order of the provided List - as well as the WaypointType. Each STOPOVER-waypoint is passed between the starting location and the destination, which are the first and last item of the waypoint list respectively.

Waypoint waypoint1 = new Waypoint(createRandomGeoCoordinatesInViewport());
Waypoint waypoint2 = new Waypoint(createRandomGeoCoordinatesInViewport());
List<Waypoint> waypoints = new ArrayList<>(Arrays.asList(new Waypoint(startGeoCoordinates),
        waypoint1, waypoint2, new Waypoint(destinationGeoCoordinates)));

routingEngine.calculateRoute(
        waypoints,
        new CarOptions(),
        new CalculateRouteCallback() {
            @Override
            public void onRouteCalculated(@Nullable RoutingError routingError, @Nullable List<Route> routes) {
                if (routingError == null) {
                    Route route = routes.get(0);
                    showRouteDetails(route);
                    showRouteOnMap(route);
                    logRouteViolations(route);
                } else {
                    showDialog("Error while calculating a route:", routingError.toString());
                }
            }
        });

By adding two additional STOPOVER-waypoints to the route using the code snippet above, we now have three route sections between the starting point and the destination as seen in the illustration below.

Additional information on the route - such as the estimated time it takes to travel to the destination and the total length of the route in meters - can be retrieved from the Route object as shown below:

long estimatedTravelTimeInSeconds = route.getDuration().getSeconds();
int lengthInMeters = route.getLengthInMeters();

Travel time and length are also available for each Section. Without additional STOPOVER-waypoints, the route will contain only one Section. If additional STOPOVER-waypoints are provided, the route is separated into several route sections between each waypoint, as well as from the starting point to the first waypoint and from the last waypoint to the destination.

Each Section contains the shape of the route in form of a GeoPolyline - represented as an array of coordinates where the first coordinate marks the starting point and the last one the destination.

This can be useful to visualize the route on the map by using, for example, map polylines in a different color for each Section. However, you can also get the polyline directly from the Route object. Below is a code snippet that shows how this could be implemented by using a MapPolyline that is drawn between each waypoint including the starting point and the destination:

GeoPolyline routeGeoPolyline = route.getGeometry();
float widthInPixels = 20;
MapPolyline routeMapPolyline = new MapPolyline(routeGeoPolyline,
        widthInPixels,
        Color.valueOf(0, 0.56f, 0.54f, 0.63f)); // RGBA

mapView.getMapScene().addMapPolyline(routeMapPolyline);

The first screenshot below shows a route without additional waypoints - and therefore only one route section. Starting point and destination are indicated by green-circled map marker objects. Note that the code for drawing the circled objects is not shown here, but can be seen from the example's source code, if you are interested.

The second screenshot shows the same route as above, but with two additional STOPOVER-waypoints, indicated by red-circled map marker objects. The route therefore, contains three route sections.

Zoom to the Route

For some use cases, it may be useful to zoom to the calculated route. The camera class provides a convenient method to adjust the viewport so that a route fits in:

GeoBox routeGeoBox = route.getBoundingBox();
// Set null values to keep the default map orientation.
camera.lookAt(routeGeoBox, new GeoOrientationUpdate(null, null));

Here we use the enclosing bounding box of the route object. This can be used to instantly update the camera: Zoom level and target point of the camera will be changed, so that the given bounding rectangle fits exactly into the viewport. Additionally, we can specify an orientation to specify more camera parameters - here we keep the default values. Note that calling lookAt() will instantly change the view.

For most use cases, a better user experience is to zoom to the route with an animation. Below you can see an example that zooms to a GeoBox plus an additional padding of 50 pixels:

private void animateToRoute(Route route) {
    // The animation should result in an untilted and unrotated map.
    double bearing = 0;
    double tilt = 0;
    // We want to show the route fitting in the map view with an additional padding of 50 pixels
    Point2D origin = new Point2D(50, 50);
    Size2D sizeInPixels = new Size2D(mapView.getWidth() - 100, mapView.getHeight() - 100);
    Rectangle2D mapViewport = new Rectangle2D(origin, sizeInPixels);

    // Animate to the route within a duration of 3 seconds.
    MapCameraUpdate update = MapCameraUpdateFactory.lookAt(
            route.getBoundingBox(),
            new GeoOrientationUpdate(bearing, tilt),
            mapViewport);
    MapCameraAnimation animation =
            MapCameraAnimationFactory.createAnimation(update, Duration.ofMillis(3000), EasingFunction.IN_CUBIC);
    mapView.getCamera().startAnimation(animation);
}

The CameraKeyframeTracks example app shows how this can look like.

Get Maneuver Instructions

Each Section contains the maneuver instructions a user may need to follow to reach the destination. For each turn, a Maneuver object contains an action and the location where the maneuver must be taken. The action may indicate actions like "depart" or directions such as "turn left".

List<Section> sections = route.getSections();
for (Section section : sections) {
    logManeuverInstructions(section);
}

And here is the code to access the maneuver instructions per section:

private void logManeuverInstructions(Section section) {
    Log.d(TAG, "Log maneuver instructions per route section:");
    List<Maneuver> maneuverInstructions = section.getManeuvers();
    for (Maneuver maneuverInstruction : maneuverInstructions) {
        ManeuverAction maneuverAction = maneuverInstruction.getAction();
        GeoCoordinates maneuverLocation = maneuverInstruction.getCoordinates();
        String maneuverInfo = maneuverInstruction.getText()
                + ", Action: " + maneuverAction.name()
                + ", Location: " + maneuverLocation.toString();
        Log.d(TAG, maneuverInfo);
    }
}

This may be useful to easily build maneuver instructions lists describing the whole route in written form. For example, the ManeuverAction enum can be used to build your own unique routing experience.

In the API Reference you can find an overview of the available maneuver actions.

Find Traffic Along a Route

You can get the overall time you are stuck in a traffic jam with:

long estimatedTrafficDelayInSeconds = route.getTrafficDelay().getSeconds();

More granular information about the traffic situation is available for the individual sections of a route: In addition to maneuvers (see above), each Section of a Route contains information of the traffic flow situation at the time when the route was calculated.

Each Section can contain a various amount of TrafficSpeed instances. These are valid along the Span until the next Span. Each Span geometry is represented by a polyline that is part of the full route's polyline shape.

The following code snippet shows how to get a TrafficSpeed element of the first Span of a Section:

Section firstSection = route.getSections().get(0);
TrafficSpeed firstTrafficSpeed = firstSection.getSpans().get(0).getTrafficSpeed();

TrafficSpeed contains the baseSpeedInMetersPerSecond, which is the expected default travel speed. Note that this may not be the same as the current speed limit on a road - as a bad road condition may justify a slower travel speed. In addition, you can get the estimated actual travel speed based on the current traffic conditions with trafficSpeedInMetersPerSecond.

Similar to the color encoding used for the traffic flow layer, you can indicate the traffic along a route using a jamFactor that has a range from 0 (no traffic) to 10 (road is blocked). See the Traffic section for more details on the traffic features of the HERE SDK.

An example how this value can be mapped to a suitable color is shown below:

Usually, the jamFactor can be interpreted like this:

• 0 <= jamFactor < 4: No or light traffic.
• 4 <= jamFactor < 8: Moderate or slow traffic.
• 8 <= jamFactor < 10: Severe traffic.
• jamFactor = 10: No traffic, ie. the road is blocked.

Note that the jamFactor indicating TrafficSpeed is calculated linear from the ratio of trafficSpeedInMetersPerSecond / baseSpeedInMetersPerSecond - without taking road types and other parameters into account. Therefore, the provided jamFactor does not necessarily match exactly the traffic flow visualization on the map view (if enabled).

If you want to visualize the traffic along a route, consider to render multiple colored MapPolyline objects for each span of a section:

• In addition, you can query traffic incidents along the route with the TrafficEngine. For this, you need to create a GeoCorridor out from the coordinates of a Route. Make sure to also specify a corridor radius by setting halfWidthInMeters. Note that the GeoCorridor is limited in length and width. Check the traffic example app to see how to use the TrafficEngine. Check also the Traffic section in this guide.

• If you are only interested in basic traffic incident information along the route, you can get a summary of the traffic situation directly from the Route instance - per Section, by calling Section.getTrafficIncidents(). The resulting TrafficIncidentOnRoute object contains a small subset of the data that is available with the TrafficEngine, but it can be useful to give a first overview.

Find Upcoming Speed Limits and More

The Route object exposes detailed information along a Route to know upcoming speed limits, street attributes and road names, dynamic traffic info and much more.

Take a look at the following attributes: Span.sectionPolylineOffset, Span.dynamicSpeedInfo, Span.streetAttributes, Span.carAttributes, Span.truckAttributes, Span.scooterAttributes, Span.walkAttributes, Span.durationInSeconds, Span.streetNames, Span.routeNumbers, Span.speedLimitInMetersPerSecond, Span.consumptionInKilowattHours, Span.functionalRoadClass, Span.duration, Span.baseDuration. Consult the API Reference for a full list of attributes.

Each attribute is given per span and it is valid for the entire length of a span. Sometimes you will also find segment IDs that indicate a portion of the road network between two intersections. A Span is a route-related concept - it is a portion of the route which has the same attributes. There can be multiple spans on the same segment and each segment usually has a start offset and an end offset.

A Span defines the smallest part of a route segment and its curvature is exposed as a list of GeoCoordinates.

Get Toll Costs Along a Route

You can get possible toll costs along the individual sections of a route.

The RouteOptions.enableTolls flag must be set to get toll costs. It is set to false by default. When this flag is enabled, toll data is requested for toll applicable transport modes as defined in the API Reference.

Get toll costs via Section.getTolls() which provides a PaymentMethod, TollFare & Toll information.

During navigation - while following a route - you can use a RoadAttributesListener that informs on new road attributes including isTollway to know if the current road may require to pay a toll.

Avoid Road Features

Use AvoidanceOptions to exclude tunnels, highways, emission zones, areas, ferries, toll costs and other road features, like harsh turns. You can pass a list of RoadFeatures, ZoneCategory, GeoBox, CountryCode elements and more.

RoadFeatures contains road specific features like tunnels or toll roads. For example, to exclude all highways from your journey, add CONTROLLED_ACCESS_HIGHWAY to the list of features.

AvoidanceOptions can be set together with other route options including localization and unit options to the options for the desired transport mode. For example, for car, use CarOptions, for trucks, TruckOptions and so on. These options are then read when calculating a route and in case of AvoidanceOptions, the engine will try to avoid the listed features.

Refresh Routes

The traffic flow information contained in a Route object is valid for the time when the route was calculated, see above. If you want to update this information at any later point in time, you can refresh the route.

It is also possible to refresh the route options for a route:

RefreshRouteOptions refreshRouteOptions = new RefreshRouteOptions(taxiOptions);

// Update the route options and set a new start point on the route.
// The new starting point must be on or very close to the original route, preferrably, use a map-matched waypoint if possible.
// Note: A routeHandle is only available, when RouteOptions.enableRouteHandle() was set to true when the original route was calculated.
routingEngine.refreshRoute(route.getRouteHandle(), mapMatchedWaypoint, refreshRouteOptions, new CalculateRouteCallback() {
    @Override
    public void onRouteCalculated(@Nullable RoutingError routingError, @Nullable List<Route> list) {
        if (routingError == null) {
            Route newRoute = list.get(0);
            // ...
        } else {
            // Handle error.
        }
    }
});

For this you need to know the RouteHandle, which must be requested via RouteOptions.enableRouteHandle() before the route was calculated.

In addition, refreshing a route can be useful to convert the route options from one transport type to another or to update specific options. If the conversion is not possible - for example, when a pedestrian route is converted to a truck route, then a routingError indicates this.

You can also shorten the route, by specifying a new starting point. The new starting point must be very close to the original route as no new route is calculated. If the new starting point is too far away, then a routingError occurs.

Note that refreshing a route is not enough when a driver deviates from a route during an ongoing turn-by-turn navigation - as the detour part needs a new route calculation. In such a case, it may be more useful to recalculate the whole route by setting a new starting point - or to use the returnToRoute() method that allows to keep the originally chosen route alternative.

During turn-by-turn navigation you can refresh a route in order to get the latest traffic delay updates ahead. Before the refreshed route can be set to the Navigator, the last location on the route needs to be set as new starting point. However, this can lead to unexpected results when the driver has already advanced while the refresh operation is ongoing. Either way, if you just want to update the ETA - and the new starting point lies on the existing route, you do not necessarily need to set the new route: Instead, you can just use the updated ETA including traffic delay time - as the rest of the route ahead remains unchanged. Setting a new route during guidance effectively starts a new navigation session.

As an alternative, consider to use the DynamicRoutingEngine that requires to set the last location and the current section index: This information will be used when the next dynamicRoutingEngineOptions.pollInterval is reached. As a trade-off, this does not lead to an immediate update and the route shape may change if a better route is found to bypass current traffic obstacles. Again, as with refreshRoute() you need to decide if you want to set the new route or not.

Return to a Route

The RoutingEngine allows to refresh an existing route (see above) based on the current traffic situation. The result may be a new route that is faster than the previous one. However, the new startingPoint must be very close to the original route.

If you need to handle a case where the startingPoint is farther away from the route, consider to use the returnToRoute() feature of the RoutingEngine. For example, a driver may decide to take a detour due to the local traffic situation. Calling returnToRoute() will also result in a new route, but it will try to resemble as much as possible from the original route without a costly route recalculation - if possible.

• Stopover waypoints are guaranteed to be passed by.
• Pass-through waypoints are only meant to shape the route, so there is no guarantee that the new route will honor or discard them.
• The current traffic situation is taken into account and may reshape the route.

Note that turn-by-turn navigation is only available for the Navigate Edition.

Import Routes From Other Services

You can import routes from other APIs and/or vendors via one of the various overloaded routingEngine.importRoute() methods. Note that this is not a 1:1 import feature, but rather a reconstruction.

A new Route object can be created from:

• A list of GeoCoordinates and RouteOptions. This can be useful, when the route should be imported from a different vendor.
• A RouteHandle. This can be useful to import a route from other HERE services. A possible use case can be to create a route on the HERE WeGo website or another web page that uses the HERE REST APIs and then transfer it to a mobile device to start a trip. Note that a map update on backend side or other changes in the real world can lead to an invalid handle. Therefore, it is recommended to use the handle only for a few hours. Although the RouteHandle encodes certain information, it requires an online connection to get the full route data from backend.

Any AvoidanceOptions that are applied may be discarded and reported as violations via the route's Section.getSectionNotices(). For example, if you request to avoid highways, but the provided coordinates match to a highway road, then the resulting route will still match the highway, but a notice is added to indicate that the desired avoidance option is violated.

In general, be aware that importing a route may not always reproduce exactly the same route as the original one. Below we look more into the details.

Import Routes from a List of Geographic Coordinates

To import a route from a list of GeoCoordinates with RouteOptions, a list of GeoCoordinates is needed that define the route shape. Such coordinates need to be very close to each other, or the calculation will fail. Such a list of coordinates can be extracted from a route calculated by a different vendor or it can be extracted from a GPX trace, for example.

List<Location> locations = Arrays.asList (new Location(new GeoCoordinates(52.518032,13.420632)), new Location(new GeoCoordinates(52.51772,13.42038)),
        new Location(new GeoCoordinates(52.51764,13.42062)), new Location(new GeoCoordinates(52.51754,13.42093)),
        new Location(new GeoCoordinates(52.51735,13.42155)), new Location(new GeoCoordinates(52.51719,13.42209)),
        new Location(new GeoCoordinates(52.51707,13.42248)), new Location(new GeoCoordinates(52.51695,13.42285)),
        new Location(new GeoCoordinates(52.5168, 13.42331)), new Location(new GeoCoordinates(52.51661,13.42387)),
        new Location(new GeoCoordinates(52.51648,13.42429)), new Location(new GeoCoordinates(52.51618,13.42513)),
        new Location(new GeoCoordinates(52.5161,13.42537)),  new Location(new GeoCoordinates(52.51543,13.42475)),
        new Location(new GeoCoordinates(52.51514,13.42449)), new Location(new GeoCoordinates(52.515001,13.424374)));

routingEngine.importRoute(locations, new CarOptions(), new CalculateRouteCallback() {
    @Override
    public void onRouteCalculated(@Nullable RoutingError routingError, @Nullable List<Route> list) {
        if (routingError == null) {
            Route newRoute = list.get(0);
            // ...
        } else {
            // Handle error.
        }
    }
});

For this option, you can select the overload of the importRoute() method that matches your desired transport type.

When importing a route from a list of GeoCoordinates and RouteOptions then the RoutingEngine will create the route shape as closely as possible from the provided geographic coordinates.

The list of coordinates is not unlimited: Please refer to the API Reference for the maximum number of supported items.

For best results use 1Hz GPS data, or points that have a spacing of a few meters. Very sparse data may result in an error.

Note: This is a beta release of this feature.

Import Routes from a RouteHandle

Below we take a look into the RouteHandle option. With the RouteHandle​(String handle) constructor you can create a RouteHandle from a given string handle. Such a string can be provided from other backend sources such as one of the available HERE REST APIs. Note that the string is only valid for a couple of hours.

Below you can find an example REST API call to create a route with a route handle string (replace YOUR_API_KEY with the actual key you have in use):

https://router.hereapi.com/v8/routes?apikey=-YOUR_API_KEY&origin=52.524465%2C13.382334&destination=52.525301%2C13.399844&return=polyline%2Csummary%2Cactions%2Cinstructions%2CrouteHandle&transportMode=car

You can copy the route handle string from the JSON response and use it with the HERE SDK like so:

routingEngine.importRoute(new RouteHandle("routeHandleStringFromBackend"),
                          new RefreshRouteOptions(TransportMode.CAR),
                          new CalculateRouteCallback() {
            @Override
            public void onRouteCalculated(@Nullable RoutingError routingError, @Nullable List<Route> list) {
                if (routingError == null) {
                    Route newRoute = list.get(0);
                    // ...
                } else {
                    // Handle error.
                }
            }
        });

The RouteHandle string identifies an already calculated route. Once imported successfully, a new Route object is created and provided as part of the CalculateRouteCallback for further use with the HERE SDK.

Note: This is a beta release of this feature.

Import Routes Offline

The HERE SDK does not support importing routes offline - at least not like shown above via a RouteHandle or by directly passing in a list of GeoCoordinates. This is only possible with the RoutingEngine that requires an online connection.

However, there is a workaround possible: By inserting an unlimited number of Waypoint items into the OfflineRoutingEngine a route can be calculated offline with the specified stopover waypoints to shape the route as close as possible as the original route. Note that the RoutingEngine supports only a limited number of waypoints of around 200 waypoints at most. When using the OfflineRoutingEngine there is no such limitation, but there may be performance trade-offs for longer routes. Therefore, it is recommended to pass only a subset of all route coordinates as waypoints - for example, filtered by a distance threshold.

Read on in the next chapter to learn about the OfflineRoutingEngine. Additional waypoints can be set in the same way as already shown above for the online RoutingEngine.

Offline Routing

In addition to the RoutingEngine, there is also an equivalent for offline use cases available: The OfflineRoutingEngine. It can be constructed in the same way as its online counterpart we already showed above:

private RoutingInterface routingEngine;
private RoutingEngine onlineRoutingEngine;
private OfflineRoutingEngine offlineRoutingEngine;

...

try {
    onlineRoutingEngine = new RoutingEngine();
} catch (InstantiationErrorException e) {
    throw new RuntimeException("Initialization of RoutingEngine failed: " + e.error.name());
}

try {
    // Allows to calculate routes on already downloaded or cached map data.
    offlineRoutingEngine = new OfflineRoutingEngine();
} catch (InstantiationErrorException e) {
    throw new RuntimeException("Initialization of OfflineRoutingEngine failed: " + e.error.name());
}

The OfflineRoutingEngine provides the same interfaces as the RoutingEngine, but the results may slightly differ as the results are taken from already downloaded or cached map data instead of initiating a new request to a HERE backend service.

This way the data may be, for example, older compared to the data you may receive when using the RoutingEngine. On the other hand, this class provides results faster - as no online connection is necessary.

However, the resulting routes contain only an empty list of maneuvers. If you need to get maneuvers during guidance, you can get them directly from the Navigator or VisualNavigator via the provided nextManeuverIndex. This is shown in the Navigation section.

All available interfaces of the OfflineRoutingEngine are also available in the RoutingEngine and both engines adopt the same interface. This makes it easy to switch between both engine instances as we will show below.

Below we show how to switch between OfflineRoutingEngine and RoutingEngine. For example, when you are on the go, it can happen that your connection is temporarily lost. In such a case it makes sense to calculate routes on the already cached or downloaded map data with the OfflineRoutingEngine.

To do so, first you need to check if the device has lost its connectivity. As a second step you can use the preferred engine:

// Sets the OfflineRoutingEngine as main engine when the device is not connected, otherwise this will set the
// RoutingEngine that requires connectivity.
private void setRoutingEngine() {
    if (isDeviceConnected()) {
        routingEngine = onlineRoutingEngine;
    } else {
        routingEngine = offlineRoutingEngine;
    }
}

Now, you can execute the same code on the current routingEngine instance, as shown in the previous sections above.

Note that the code for isDeviceConnected() to check if the device is online or not is left out here. You may use a 3rd party API for this or try to make an actual connection - and when that fails, you can switch to the OfflineRoutingEngine - or the other way round: You can try offline routing first to provide a fast experience for the user, but when no map data is available, you can try online.

Optimize the Waypoint Order

The OfflineEngine (see above) also allows offline waypoint sequencing to support the travelling salesman use case. Note that the feature is not yet supported for the online RoutingEngine.

Multiple stopover Waypoint items along a Route can be ordered automatically by setting the RouteOptions.optimizeWaypointsOrder: If set to true, the OfflineRoutingEngine will try to optimize the waypoints order to reach the destination faster or to make the route shorter - while ensuring that all set waypoints of type stopover are reached.

The specified OptimizationMode determines if the reordering results in a faster or shorter route.

At least 5 or more waypoints (including starting point and destination) need to be set in order to see an effect on the route calculation. If needed, the reshuffled waypoints can be identified by comparing the originalCoordinates of each RoutePlace with the user-defined coordinates. RoutePlace objects are provided for each Section of a Route object as each stopover splits the route into sections. During navigation the MilestoneStatus event informs if a user-defined waypoint has been reached.

Note that this is a beta release of this feature.

Show Reachable Area

With isoline routing you can generate polygons to represent the area of reach from a given point based on time, distance or energy consumption: The polygon will encompass all destinations that can be reached in a specific amount of time, a maximum travel distance, or even the charge level available at an electric vehicle.

Some examples of how this could be useful:

• Users can find restaurants within a 2 km walking distance from their current location.
• Users can search for hotels based on the distance to sights, for example, to find hotels within a 20 minute drive to major attractions - such as Disney World and Universal Studios in Orlando, Florida USA.

Below we show a consumption based Isoline example for an electric vehicle scenario, where the driver wants to know which points are reachable within the provided limit of 400 Wh: So, the goal is to consume 400 Wh or less - and the question is: What can the driver reach within this energy limit?

The result will be a GeoPolygon shape that you can show on the map to provide the driver a visual orientation:

List<Integer> rangeValues = Collections.singletonList(400);

IsolineOptions.Calculation calculationOptions =
        new IsolineOptions.Calculation(IsolineRangeType.CONSUMPTION_IN_WATT_HOURS, rangeValues, IsolineCalculationMode.BALANCED);
IsolineOptions isolineOptions = new IsolineOptions(calculationOptions, getEVCarOptions());

routingEngine.calculateIsoline(new Waypoint(startGeoCoordinates), isolineOptions, new CalculateIsolineCallback() {
    @Override
    public void onIsolineCalculated(RoutingError routingError, List<Isoline> list) {
        if (routingError != null) {
            showDialog("Error while calculating reachable area:", routingError.toString());
            return;
        }

        // When routingError is nil, the isolines list is guaranteed to contain at least one isoline.
        // The number of isolines matches the number of requested range values. Here we have used one range value,
        // so only one isoline object is expected.
        Isoline isoline = list.get(0);

        // If there is more than one polygon, the other polygons indicate separate areas, for example, islands, that
        // can only be reached by a ferry.
        for (GeoPolygon geoPolygon : isoline.getPolygons()) {
            // Show polygon on map.
              Color fillColor = Color.valueOf(0, 0.56f, 0.54f, 0.5f); // RGBA
            MapPolygon mapPolygon = new MapPolygon(geoPolygon, fillColor);
            mapView.getMapScene().addMapPolygon(mapPolygon);
            mapPolygons.add(mapPolygon);
        }
    }
});

This finds the area that an electric vehicle can reach by consuming 400 Wh or less, while trying to take the fastest possible route into any possible straight direction from start.

Why fastest? This depends on the route's optimization mode (we have shown in earlier sections above) - of course, you can specify any mode. Note that each isoline needs exactly one center point from where a journey may start.

Since we are interested in the energy consumption, we also provided EVCarOptions. These options must include battery specifications as shown in the electric vehicle routing section below. If you provide a time or distance limit instead as range type, you can provide the usual route options as shown earlier.

The IsolineRangeType defines the type of the provide range value. Here we use 400. You can provide multiple values - and as a result you would then get multiple Isoline objects - one for each provided range value.

On the other hand, each Isoline object can contain multiple polygons for special cases, such as when the area of reach includes an island. Such areas are provided as separate polygons.

The shape of the resulting polygon can be defined by the maxPoints parameter. It determines the number of vertices of the GeoPolygon. Note that this parameter does not influence the actual calculation of the polygon, but just its apearance.

The screenshot shows the center location of the polygon, indicated by a big green circle. It also shows a possible route beyond this area with two additional found charging stations along the route. In the next section you can find how to calculate routes for electric vehicles.

Get Routes for Electric Vehicles

Electric vehicle (EV) usage and sales continue growing around the world. How can HERE help to provide the best routes for electric vehicles?

• HERE EV Routing delivers optimized routes for EVs to get from A to B, minimizing the number of charging stops and optimizing battery charge times (based on the vehicle’s consumption model). It also takes into account a number of factors when planning trips, including topography, road geometry, real-time traffic information and traffic patterns.

• Instead of searching for charging stations along the route, HERE analyzes the charging time at every possible station and chooses the most appropriate combination that will deliver the shortest driving and charging times.

Getting EV routes is simple. Similar to getting car or truck routes, for EV routing you just need to add electric vehicle specific route options. This way you can get routes for electric vehicles in the same way as for other transport modes. Just specify EVRouteOptions and add them to the calculateRoute() method:

routingEngine.calculateRoute(waypoints, getEVCarOptions(), new CalculateRouteCallback() {
    @Override
    public void onRouteCalculated(RoutingError routingError, List<Route> list) {
        if (routingError != null) {
            showDialog("Error while calculating a route: ", routingError.toString());
            return;
        }

        // Use routes from list.
    }
});

By default, EVRouteOptions will not include the required parameters to ensure that a destination can be reached without running out of energy.

To ensure this, you need to set the required parameters that may add charging stations to the route - while still optimizing the result for overall travel time.

Below you can see an example how to create such options:

private EVCarOptions getEVCarOptions()  {
    EVCarOptions evCarOptions = new EVCarOptions();

    // The below three options are the minimum you must specify or routing will result in an error.
    evCarOptions.consumptionModel.ascentConsumptionInWattHoursPerMeter = 9;
    evCarOptions.consumptionModel.descentRecoveryInWattHoursPerMeter = 4.3;
    evCarOptions.consumptionModel.freeFlowSpeedTable = new HashMap<Integer, Double>() {{
        put(0, 0.239);
        put(27, 0.239);
        put(60, 0.196);
        put(90, 0.238);
    }};

    // Must be 0 for isoline calculation.   
    evCarOptions.routeOptions.alternatives = 0;

    // Ensure that the vehicle does not run out of energy along the way
    // and charging stations are added as additional waypoints.
    evCarOptions.ensureReachability = true;

    // The below options are required when setting the ensureReachability option to true
    // (AvoidanceOptions need to be empty).
    evCarOptions.avoidanceOptions = new AvoidanceOptions();
    evCarOptions.routeOptions.speedCapInMetersPerSecond = null;
    evCarOptions.routeOptions.optimizationMode = OptimizationMode.FASTEST;
    evCarOptions.batterySpecifications.connectorTypes =
            new ArrayList<>(Arrays.asList(ChargingConnectorType.TESLA,
                ChargingConnectorType.IEC_62196_TYPE_1_COMBO, ChargingConnectorType.IEC_62196_TYPE_2_COMBO));
    evCarOptions.batterySpecifications.totalCapacityInKilowattHours = 80.0;
    evCarOptions.batterySpecifications.initialChargeInKilowattHours = 10.0;
    evCarOptions.batterySpecifications.targetChargeInKilowattHours = 72.0;
    evCarOptions.batterySpecifications.chargingCurve = new HashMap<Double, Double>() {{
        put(0.0, 239.0);
        put(64.0, 111.0);
        put(72.0, 1.0);
    }};

    // Note: More EV options are available, the above shows only the minimum viable options.

    return evCarOptions;
}

Here we define the EVConsumptionModel to specify the energy consumption model of the electric vehicle. We also add BatterySpecifications. With these options, additional charging stations may be inserted into the route as additional stopovers - which means that the route will be split into more sections depending on the number of inserted charging stations.

All options can influence the time it takes to calculate a route for electric vehicles. Most importantly, a large battery capacity - as set with totalCapacityInKilowattHours - can decrease the need for charging stops - as well as a fully loaded battery when starting the trip - see initialChargeInKilowattHours: A relatively low value means that the route must include charging stations near the beginning. Otherwise, the route calculation may even fail.

Currently, the engine will only include charging stations that are in service. Charging stations that our out-of-service or in maintenance are not considered. However, due to the dynamic situation at charging stations, the engine does not consider if a station is currently occupied or reserved. So, in worst case, it can happen that you arrive at a station and it is currently being used to charge another car.

Note that we also specify the available battery connector types, so that charging stations that do not provide a compatible connector to charge the car will be excluded.

Especially for longer journeys with electric vehicles, it is important to plan for charging stops along the way. After all, charging stations are much less common than gas stations. With the above options, the RoutingEngine tries to find the fastest route, i.e., one with the lowest overall time consumed to reach the destination, while ensuring that the vehicle does not run out of energy along the way.

The result of the calculation is a route optimized for electric vehicles - instead of just adding any charging stations found along the way - as we have shown in the Search Along a Route section.

Once the route is calculated, you can gather more useful information. The code snippet shown below will log the estimated energy consumption per Section and list the actions you need to take in order to charge the battery - if needed:

// Find inserted charging stations that are required for this route.
// Note that this example assumes only one start waypoint and one destination waypoint.
// By default, each route has one section.
int additionalSectionCount = route.getSections().size() - 1;
if (additionalSectionCount > 0) {
    // Each additional waypoint splits the route into two sections.
    Log.d("EVDetails", "Number of required stops at charging stations: " + additionalSectionCount);
} else {
    Log.d("EVDetails","Based on the provided options, the destination can be reached without a stop at a charging station.");
}

int sectionIndex = 0;
List<Section> sections = route.getSections();
for (Section section : sections) {
    EVDetails evDetails = section.getEvDetails();
    if (evDetails == null) {
        return;
    }    
    Log.d("EVDetails", "Estimated net energy consumption in kWh for this section: " + evDetails.consumptionInKilowattHour);
    for (PostAction postAction : section.getPostActions()) {
        switch (postAction.action) {
            case CHARGING_SETUP:
            Log.d("EVDetails", "At the end of this section you need to setup charging for " + postAction.duration.getSeconds() + " s.");
                break;
            case CHARGING:
            Log.d("EVDetails", "At the end of this section you need to charge for " + postAction.duration.getSeconds() + " s.");
                break;
            case WAIT:
            Log.d("EVDetails", "At the end of this section you need to wait for " + postAction.duration.getSeconds() + " s.");
                break;
            default: throw new RuntimeException("Unknown post action type.");
        }
    }

    Log.d("EVDetails", "Section " + sectionIndex + ": Estimated battery charge in kWh when leaving the departure place: " + section.getDeparturePlace().chargeInKilowattHours);
    Log.d("EVDetails", "Section " + sectionIndex + ": Estimated battery charge in kWh when leaving the arrival place: " + section.getArrivalPlace().chargeInKilowattHours);

    // Only charging stations that are needed to reach the destination are listed below.
    ChargingStation depStation = section.getDeparturePlace().chargingStation;
    if (depStation != null  && depStation.id != null && !chargingStationsIDs.contains(depStation.id)) {
        Log.d("EVDetails", "Section " + sectionIndex + ", name of charging station: " + depStation.name);
        chargingStationsIDs.add(depStation.id);
        addCircleMapMarker(section.getDeparturePlace().mapMatchedCoordinates, R.drawable.required_charging);
    }

    ChargingStation arrStation = section.getDeparturePlace().chargingStation;
    if (arrStation != null && arrStation.id != null && !chargingStationsIDs.contains(arrStation.id)) {
        Log.d("EVDetails", "Section " + sectionIndex + ", name of charging station: " + arrStation.name);
        chargingStationsIDs.add(arrStation.id);
        addCircleMapMarker(section.getArrivalPlace().mapMatchedCoordinates, R.drawable.required_charging);
    }

    sectionIndex += 1;
}

Note that postAction.duration.getSeconds() provides the estimated time it takes to charge the battery. This time is included in the overall route calculation and the estimated time of arrival (ETA).

Below is a screenshot of how a resulting route may look.

Here you can see that the route required two stops at a charging station - indicated by red markers. The route contains three sections as each charging station splits the route - when inserting additional waypoints.

The first section includes a post action which describes the charging stop. It contains the information on the expected arrival charge among other information.

Unless specified otherwise, the energy consumption is assumed to be in Wh.

Define the Consumption Model

The following parameters define a consumption model to get more accurate results for electric vehicles:

• ascentConsumptionInWattHoursPerMeter: Rate of energy consumed per meter rise in elevation.
• descentRecoveryInWattHoursPerMeter: Rate of energy recovered per meter fall in elevation.
• freeFlowSpeedTable: Function curve specifying consumption rate at a given free flow speed on a flat stretch of road.
• trafficSpeedTable: Function curve specifying consumption rate at a given speed under traffic conditions on a flat stretch of road.
• auxiliaryConsumptionInWattHoursPerSecond: Rate of energy consumed by the vehicle's auxiliary systems (e.g., air conditioning, lights) per second of travel.

A consumption speed table defines the energy consumption rate when the vehicle travels on a straight road without elevation change at a given speed in km/h. It represents a piecewise linear function.

Here is an example of a free flow speed list. On the left you see the speed and on the right, consumption:

• 0: 0.239
• 27: 0.239
• 45: 0.259
• 60: 0.196
• 75: 0.207
• 90: 0.238
• 100: 0.26
• 110: 0.296
• 120: 0.337
• 130: 0.351

In a graph it will look like this:

You can specify two different consumption speed tables - free flow speed table and traffic speed tables:

• Free flow speed: Describes the energy consumption when travelling at constant speed.
• Traffic speed: Describes the energy consumption when travelling under heavy traffic conditions, for example, when the vehicle is expected to often change the travel speed at a given average speed.

If a trafficSpeedTable is not provided then only the freeFlowSpeedTable is used for calculating speed-related energy consumption.