- Overview
- Contributing
- License
- Additional Resources
Welcome to the Monitoring School Zone Speed project. This example uses the iPhone 11 Location Services to simulate a GPS device. Polylines are used to identify school zones where the speed limit is 25 MPH. When the device enters a school zone its speed is monitored. If the speed exceeds 30 MPH for five seconds a warning e-mail message is sent to the owner of the device.
The following is a list of learning objectives for the Monitoring School Zone Speed project:
- How to surface sensor data for an iPhone 11.
- How to use a socket connection to publish data to ESP.
- How to use polylines as a trip wire or boundary
The iPhone 11 includes a Global Positioning System (GPS) chip that functions like the chip in stand-alone GPS devices. Because GPS signals are not always strong (e.g., inside building), signals from cellular towers and Wi-Fi networks are also used in a system called assisted GPS. This allows iPhone 11 Location Services to determine the location of the device quickly and accurately.
In addition to geographic location, Location Services determine the altitude and orientation of the device. Location Services uses all the available components on the device including:
- GPS
- Wi-Fi
- Bluetooth
- Magnetometer
- Barometer
- Cellular Hardware
The following table lists the Location Services data items available and provides a brief description of each:
| Data Item | Description |
|---|---|
| Datetime | The date and time the location was determined in UNIX EPOCH time. |
| Latitude | The latitude of the device in degrees. |
| Longitude | The longitude of the device in degrees. |
| Altitude | The altitude of the device in meters. |
| Speed | The speed at which the device is travelling in meters per second. |
| Course | The direction in which the device is moving in degrees. |
| Floor | The floor of the building on which the device is located if floor information is available. |
| Horizontal accuracy | The radius of uncertainty for the location, measured in meters. A negative value indicates that the latitude and longitude are invalid. |
| Vertical accuracy | A zero or positive number indicates the value in the altitude property is plus or minus the specified number of meters. A negative value indicates that the altitude value is invalid. |
SensorLog is an iPhone app developed by Dr. Bernd Thomas that reads and collects data from the sensors on the device. It can write the information to a file (csv or JSON) or stream the data by way of a socket connection to a user-specified server.
SensorLog reads the following types of sensor information. You can select which items to include:
- Device orientation
- Battery level
- Position (i.e., location)
- Heading
- Accelerometer
- Gyroscope
- Device motion
- Activity
- Pedometer
- Altimeter
- Audio level
Figure 1 - SensorLog Home Screen
NOTE: Location Services must be turned on and SensorLog must have permission to access the device’s location. Refer to the Apple Support site for instructions.
The SensorLog home screen provides current information and buttons to control the app. The top of the screen displays the current location of the device on an embedded Google Maps. Under the map is the device’s current location values as provided by Location Services.
The current configuration is displayed near the bottom of the screen. Finally, there is a menu bar at the bottom that includes a Start/Stop button and a button to access the configuration.
Figure 2 - SensorLog Configuration Screen Page 1
The Configuration screen is accessed by way of the Configuration button on the SensorLog home screen. This is where you configure how SensorLog surfaces the sensor data.
The Logging Settings are found at the top of the Configuration screen. This is where you set the logging format and rate, and where you select the log to file or log to stream toggle.
If log to stream is selected, you will need to configure the Streaming Settings. Refer to the section, Configure SensorLog to Stream Information, for more information.
Figure 3 - SensorLog Configuration Screen Page 2
You can access the Sensors and Data section of Configuration by scrolling down. This is where you select which data items to include in the stream or file.
This project uses a socket connection to publish events from SensorLog on the iPhone 11 device to the inputGPS source window in SAS ESP. When the ESP connector implements a socket connection, it might act as a client or server. When the connector specifies only the port number (e.g., :55555) it implements the server and the network peer (SensorLog) is the client.
Figure 4 - Socket Connection
The server (ESP) includes a socket that is bound to a specific port number. When the project is running ESP is listening to the socket for a connection request from a client.
The client (SensorLog) knows the IP Address and port number on which the server is listening. It initiates a connection request which is accepted by the server. The server makes the connection and events can now be published from the client to the server.
The following shows the connector configuration for a socket connector acting as a server:
Figure 5 - Socket Connector Configuration
Figure 6 - Geofence
A geofence is a virtual perimeter around a real-world geographic area. It has two input sources, position and geometries. The position source window streams the geographic positions of one or more objects. The geometries source window provides definitions for areas of interest, or geometries.
The geofence window functions like a left outer join with the position source being the left side of the join and the geometries source acting as the right side. Position data is always output. However, geometry data is output only when the position is within the perimeter defined by the geofence.
There are three types of geometries used to form a geofence:
- Circle – Set of coordinates (center point) surrounded by a radius.
- Polygon – Multiple sets of coordinates forming a closed ring.
- Polyline – Multiple sets of coordinates forming a trip wire or boundary.
A polyline is a shape that defines a border or trip wire. It is comprised of multiple sets of coordinates that form one or more segments. The following figure shows the components when processing a polyline:
Figure 7 - Polyline Components
The above polyline includes four points, P1 through P4. The order of the coordinate sets in the polyline definition determines the order of the points. Point P1 is considered the origin of the polyline. Because there are four points, there are three segments. Segment 1 is formed by connecting points P1 and P2, segment 2 by connecting P2 and P3, and segment 3 by connecting P3 and P4.
The polyline has a radius defined of 50 meters. If polyline strict projection is enabled, a perpendicular line is drawn at points P1 and P4 and the portion of the radius outside the line is excluded. If any position is within the radius (not counting for strict projection) the Geometry ID and all other output fields are included in the output event.
When a position is within the radius, the distance to the closest segment is calculated. If the position is to the left of the polyline the distance value is negative. If to the right the value is positive. This allows you to detect to which side of the polyline the position lies. The distance to the origin (i.e., P1) is also calculated.
This project uses polylines to define school zones. This is accomplished by determining if the position is within the radius, the closest segment of the polyline, and whether the distance value is positive or negative.
The following figure shows several examples of vehicles near or in the school zone. The table that follows the figure shows the results of evaluating whether the vehicle is in the school zone. Polyline strict projection is enabled.
Figure 8 - Polyline Examples
| Vehicle | Geometry ID Returned | Closest Segment | Distance | In Speed Zone |
|---|---|---|---|---|
| A | No | N/A | N/A | No |
| B | No | N/A | N/A | No |
| C | Yes | 1 | Positive | No |
| D | Yes | 3 | Negative | No |
| E | Yes | 1 | Negative | Yes |
| F | Yes | 2 | Negative | Yes |
| G | Yes | 3 | Positive | Yes |
Here is the project in ESP Studio.
Figure 9 - ESP Project
The inputGPS source window inserts the iPhone Location Services data. This is the SensorLog data in csv format by way of a socket connection or file. The window includes two input data connectors, one for the socket connection and one for the csv file. You can select which input data connector to use.
The following shows the schema for the inputGPS source window:
Figure 10 - inputGPS Source Window Schema
The convertToImperial compute window creates several new fields using expressions. The window passes through the id, locDatetime, vendorID, locLatitude, and locLongitude fields. The following table shows the created fields and the expressions used to populate them:
| New Field | Expression |
|---|---|
| speedMPH | locSpeed * 2.23694 |
| compassDir | if (locCourse >= 315) or (locCourse < 45) then return "N" else if (locCourse >= 45) and (locCourse < 135) then return "E" else if (locCourse >= 135) and (locCourse < 225) then return "S" else return "W" |
| altitudeFt | locAltitude * 3.28084 |
| locationError | if locHorAcc >= 0 and locHorAcc <= 20 and locVerAcc >= 0 and locVerAcc <= 20 then return 0 else return 1 |
The field locationError is used to control splitting the output. There are two output slots, one as input to the monitorSchoolZone geofence window (slot 0) and one to the errors copy window (slot 1). This removes any events with invalid or inaccurate values from the normal flow of the project.
The errors copy window receives any events from the convertToImperial window where the horizontal or vertical accuracy are out of tolerance. The window retains the most recent hour of error events.
The schoolZones source window reads the polyline definitions from the school_zone2.csv file. The output of this window is used as the geometries input to the monitorSchoolZones geofence window.
The following shows the schema for the schoolZones source window:
Figure 11 - schoolZones Source Window Schema
The monitorSchoolZones geofence window receives positional input from slot 0 of the convertToImperial compute window. It receives geometric information from the schoolZones source window. Polygon strict projection is enabled.
The following figure shows the output map for the monitorSchoolZones window. Whenever a position is within the defined radius of 50 meters, the Geometry ID and all other output fields are populated:
Figure 12 - monitorSchoolZones Geofence Window Output Map
The inSchoolZone filter window passes through events when the position is within a defined school zone . All other events are filtered. The following is the expression used to pass through events when the position is in a school zone:
geoid != '' AND
(
(closest == 1 and geodist < 0) OR
(closest == 2) OR
(closest == 3 and geodist > 0)
)
The changeToInsert window receives events for the inSchoolZone filter window, converts them to inserts, and outputs the events to the violations pattern window. This window is required because a filter window can output inserts, updates, deletes, and upserts and a pattern window requires inserts only.
The violations Pattern window is used to detect speeding violations (speed in excess of 30 MPH) when a vehicle is in a school zone. The pattern is looking for five unique events where the speed is greater than 30 MPH.
The following figure shows the detectSpeeding pattern used by the violations pattern window:
Figure 13 - violations Pattern Window detectSpeeding Pattern
The detectSpeeding pattern uses vendorID as the index field. This means the vendorID must be the same in all events of interest.
Each of the five events of interest (EOI) use the same expression, (speedMPH > 30). The logic is as follows:
fby(is(e1),is(e2),is(e3),is(e4),is(e5))
The fby operator requires the first operand to be true, followed by the second operand, and so on. The EOIs must occur in sequence and each EOI must have a unique event applied to it. The is operator requires the EOI to be true and then fixes it so that no other events are applied to the EOI.
What this all means is that for the pattern to be true, there must be five unique events over 30 MPH in sequence.
The sendWarning Notification window delivers a warning e-mail message when a school zone speeding violation occurs. For this project, the messages are written only to the server console. The throttle value for the window is set to send one message per minute, regardless of the number of input events.
For additional information about Notification windows, refer to the Using Notification Windows section of the SAS ESP 6.2 documentation.
- SAS Event Stream Processing 6.2
- SensorLog app for iPhone 11
All files required to execute the project are available in the files directory of this page. Download the following files:
- school_zones.csv – Polyline definitions for school zones
- gps_input.csv – Input data file recorded from SensorLog
- sensor_log.xml – ESP project
- run_polyline.sh – Startup script to execute ESP project
- dashboard.xml – ESP Streamviewer dashboard
Create a server directory for the project’s files. The following is an example of the command to create a directory:
mkdir /home/zestem/polyline
Upload all the files to the directory you created. Change the permission of the run_polyline.sh startup script so it is executable. The following is an example of the command to do this:
chmod 777 /home/zestem/ployline/run_polyline.sh
This project includes a project file, sensor_log.xml, you can execute. The project includes two input data connectors, one to read a csv file and the other to use in a socket connection. Instructions for both are included in the following sections.
- Execute the project from the Command Line
- Execute the project with SAS ESP Studio
- Making a Socket Connection
Execute the project at the Command Line
Execute the project with SAS ESP Studio
Making a Socket Connection
We welcome your contributions! Please read CONTRIBUTING.md for details on how to submit contributions to this project.
This project is licensed under the Apache 2.0 License.