The Parking Lot Manager Demo is a skill that learns a set of parking spaces and then monitors the status of those spaces.  The user is initially guided through parking lot configuration via a web interface.  They will define the learning parameters which should be used during “learning mode” (example: camera region of interest and identifying parking spaces of interest).  Once configured, the skill will learn the parking lot over a user defined duration (recommended 24 hours or more such that all spots will have been occupied at some point). Once the learning operation is complete, the user will verify that the spaces of interest have been identified correctly and then initiate monitoring mode, at which point parking lot occupancy levels will be reported via the portal.
 
Want to manage one or more full parking lots using multiple cameras? Let Lexmark help you gain the visibility you want by integrating an array of parking lot cameras into one cohesive view.

Lexmark

Parking Lot Manager Demo

The Person Occupancy skill monitors live camera feed(s) and quantifies the number of people in the camera frame over time.

Cameras should be set up via the portal on the device that will be running the skill. This skill requires 1 to 4 Real Time Streaming Protocol (RTSP) networked cameras.

When loading the Person Occupancy skill into a workflow, the user may set:
1. “Optra Send Frequency” – The frequency in seconds that the Optra will report information back to the portal. Defaulted to 60 seconds.
2. “Camera Frames per Second (FPS) Limiter” – The skill will attempt to run at this FPS, but not higher. If the FPS limiter value is higher than the skill/Optra can maintain, the actual FPS may be a lower value.  The actual FPS is reported in the skill logs.  Defaulted to 7 FPS.
3. “Confidence Threshold” – This is the confidence threshold used to determine if a detected object is a person or not.  A lower threshold will capture more  people, but have a higher false positive rate.  A higher threshold will be more specific when identifying an object as a person, but may not detect difficult to see individuals.  The highest threshold is 0.99, while the lowest threshold is 0.01.  Defaulted to 0.5.

Data that the Person Occupancy can report via the portal:
1. “People Average Occupancy Camera 1” – The average occupancy reported from camera 1.
2. “People Average Occupancy Camera 2” – The average occupancy reported from camera 2.
3. “People Average Occupancy Camera 3” – The average occupancy reported from camera 3.
4. “People Average Occupancy Camera 4” – The average occupancy reported from camera 4.
5. “People Maximum Occupancy Camera 1” – The maximum occupancy reported from camera 1.
6. “People Maximum Occupancy Camera 2” – The maximum occupancy reported from camera 2.
7. “People Maximum Occupancy Camera 3” – The maximum occupancy reported from camera 3.
8. “People Maximum Occupancy Camera 4” – The maximum occupancy reported from camera 4.
9. “Created At” – The date/time stamp of the reported data.
10. “Device ID” – A unique numeric identifier of the Optra device reporting.
11. “Skill ID” – A unique numeric identifier of the skill instance reporting.
12. “Serial Number” –  The serial number of the Optra device reporting.
13. “Device Name” – The human readable name of the Optra device reporting as assigned in the Portal.
14. “Latitude” – The latitude of the Optra device reporting.
15. “Longitude” – The longitude of the Optra device reporting.

Person Occupancy

Pair a network camera with this Optra skill to monitor product assembly and insure the products you build every day meet the quality standards you aspire to.  The Manufacturing Vision skill is designed to integrate with your MES (Manufacturing Execution System) via a REST API hosted by the Optra Device.  Allow Lexmark to help you build custom computer vision models to inspect critical elements of your products.  

The Manufacturing Vision application is designed to operate “on the edge,” minimizing the risk of assembly downtime due to loss of internet connectivity, and optionally enabling an air-gapped network architecture.

The machine learning model used by the skill can be updated or changed via the REST API and the business logic is left to the user’s MES, making the Manufacturing Vision skill highly customizable to the user’s application and system needs.

The REST API supports secure HTTPS access only, meaning a username and password (or authentication token) are required for each request. These login credentials are supplied in the Auth Header using Basic Authentication.

Manufacturing Vision

Automated visual inspection for manufacturing.

The Vehicle Monitor skill monitors live camera feed(s) and visualizes the number of vehicles in the camera frame over time.

Cameras should be set up via the portal on the device that will be running the skill.  This skill requires 1 to 4 Real Time Streaming Protocol (RTSP) networked cameras.

When loading the Vehicle Monitor skill into a workflow, the user may set:
1. “Optra Send Frequency” – The frequency in seconds that the Optra will report information back to the portal.  Defaulted to 60 seconds.
2. “Camera Frames per Second (FPS) Limiter” – The skill will attempt to run at this FPS, but not higher.  If the FPS limiter value is higher than the skill/Optra can maintain, the actual FPS may be a lower value.  The actual FPS is reported in the skill logs.  Defaulted to 7 FPS.
3. “Confidence Threshold” – This is the confidence threshold used to determine if a detected object is a vehicle or not.  A lower threshold will capture more vehicles but have a higher false positive rate.  A higher threshold will be more specific when identifying an object as a vehicle may not detect vehicles that are difficult to see.  The highest threshold is 0.99, while the lowest threshold is 0.01.  Defaulted to 0.5

Data that the Vehicle Monitor can report via the portal:
1. “Vehicle Average Occupancy Camera 1” – The average occupancy reported from camera 1.
2. “Vehicle Average Occupancy Camera 2” – The average occupancy reported from camera 2.
3. “Vehicle Average Occupancy Camera 3” – The average occupancy reported from camera 3.
4. “Vehicle Average Occupancy Camera 4” – The average occupancy reported from camera 4.
5. “Vehicle Maximum Occupancy Camera 1” – The maximum occupancy reported from camera 1.
6. “Vehicle Maximum Occupancy Camera 2” – The maximum occupancy reported from camera 2.
7. “Vehicle Maximum Occupancy Camera 3” – The maximum occupancy reported from camera 3.
8. “Vehicle Maximum Occupancy Camera 4” – The maximum occupancy reported from camera 4
9. “Created At” – The date/time stamp of the reported data.
10. “Device ID” – A unique numeric identifier of the Optra device reporting.
11. “Skill ID” – A unique numeric identifier of the skill instance reporting.
12. “Serial Number” –  The serial number of the Optra device reporting.
13. “Device Name” – The human readable name of the Optra device reporting as assigned in the Portal.
14. “Latitude” – The latitude of the Optra device reporting.
15. “Longitude” – The longitude of the Optra device reporting.

Vehicle Monitor

This utility can be highly useful for periodically logging the internet speed of an Optra Edge device’s network.  The skill will run speedtest-cli at the frequency specified within the skill configuration page and will report the results via the Optra Portal portal.  The speed test outputs from the skill may directed to the user’s “Action” of choice within their workflow.  See speedtest-cli at:  https://pypi.org/project/speedtest-cli/

speedtest-cli

This skill is a wrapper around the speedtest utility.

The Vehicle Monitor Plus skill monitors live camera feed(s) and quantifies the number of vehicles in the camera frame over time.  The Plus version of the skill enables users to see activity heatmaps for various time periods, enabling you to understand the vehicle activity patterns within each camera frame.

Cameras should be set up via the portal on the device that will be running the skill.  This skill requires 1 to 4 Real Time Streaming Protocol (RTSP) networked cameras.

When loading the Vehicle Monitor Plus skill into a workflow, the user may set:
1. “Optra Send Frequency” – The frequency in seconds that the Optra will report information back to the portal. Defaulted to 60 seconds
2. “Camera Frames per Second (FPS) Limiter” – The skill will attempt to run at this FPS, but not higher.  If the FPS limiter value is higher than the skill/Optra can maintain, the actual FPS may be a lower value.  The actual FPS is reported in the skill logs.  Defaulted to 7 FPS.
3. “Confidence Threshold” – This is the confidence threshold used to determine if a detected object is a vehicle or not.  A lower threshold will capture more vehicles but have a higher false positive rate.  A higher threshold will be more specific when identifying an object as a vehicle but may not detect vehicles that are difficult to see.  The highest threshold is 0.99, while the lowest threshold is 0.01.  Defaulted to 0.5

Vehicle Monitor Plus offers a web interface that allows the user to view a historical heatmap of vehicle activity.

Data that the Vehicle Monitor Plus can report via the portal:
1. “Vehicle Average Occupancy Camera 1” – The average occupancy reported from camera 1.
2. “Vehicle Average Occupancy Camera 2” – The average occupancy reported from camera 2.
3. “Vehicle Average Occupancy Camera 3” – The average occupancy reported from camera 3.
4. “Vehicle Average Occupancy Camera 4” – The average occupancy reported from camera 4.
5. “Vehicle Maximum Occupancy Camera 1” – The maximum occupancy reported from camera 1.
6. “Vehicle Maximum Occupancy Camera 2” – The maximum occupancy reported from camera 2.
7. “Vehicle Maximum Occupancy Camera 3” – The maximum occupancy reported from camera 3.
8. “Vehicle Maximum Occupancy Camera 4” – The maximum occupancy reported from camera 4.
9. “Created At” – The date/time stamp of the reported data.
10. “Device ID” – A unique numeric identifier of the Optra device reporting.
11. “Skill ID” – A unique numeric identifier of the skill instance reporting.
12. “Serial Number” –  The serial number of the Optra device reporting.
13. “Device Name” – The human readable name of the Optra device reporting as assigned in the Portal.
14. “Latitude” – The latitude of the Optra device reporting.
15. “Longitude” – The longitude of the Optra device reporting.

Vehicle Monitor Plus

Easy and secure vehicle counting and density mapping

The Person Occupancy Plus skill monitors live camera feed(s) and quantifies the number of people in the camera frame over time.  The Plus version of the skill enables users to see density heatmaps for various time periods, enabling you to understand the activity patterns within each camera frame.

Cameras should be set up via the portal on the device that will be running the skill.  This skill requires 1 to 4 Real Time Streaming Protocol (RTSP) networked cameras.

When loading the Person Occupancy Plus skill into a workflow, the user may set:
1. “Optra Send Frequency” – The frequency in seconds that the Optra will report information back to the portal.  Defaulted to 60 seconds.
2. “Camera Frames per Second (FPS) Limiter” – The skill will attempt to run at this FPS, but not higher.  If the FPS limiter value is higher than the skill/Optra can maintain, the actual FPS may be a lower value.  The actual FPS is reported in the skill logs. Defaulted to 7 FPS.
3. “Confidence Threshold” – This is the confidence threshold used to determine if a detected object is a person or not.  A lower threshold will capture more people but have a higher false positive rate.  A higher threshold will be more specific when identifying an object as a person but may not detect difficult to see individuals.  The highest threshold is 0.99, while the lowest threshold is 0.01.  Defaulted to 0.5

Person Occupancy Plus offers a web interface that allows the user to view a historical heatmap of people activity.

Data that the Person Occupancy Plus can report via the portal:
1. “People Average Occupancy Camera 1” – The average occupancy reported from camera 1.
2. “People Average Occupancy Camera 2” – The average occupancy reported from camera 2.
3. “People Average Occupancy Camera 3” – The average occupancy reported from camera 3.
4. “People Average Occupancy Camera 4” – The average occupancy reported from camera 4.
5. “People Maximum Occupancy Camera 1” – The maximum occupancy reported from camera 1.
6. “People Maximum Occupancy Camera 2” – The maximum occupancy reported from camera 2.
7. “People Maximum Occupancy Camera 3” – The maximum occupancy reported from camera 3.
8. “People Maximum Occupancy Camera 4” – The maximum occupancy reported from camera 4.
9. “Created At” – The date/time stamp of the reported data.
10. “Device ID” – A unique numeric identifier of the Optra device reporting.
11. “Skill ID” – A unique numeric identifier of the skill instance reporting.
12. “Serial Number” –  The serial number of the Optra device reporting.
13. “Device Name” – The human readable name of the Optra device reporting as assigned in the Portal.
14. “Latitude” – The latitude of the Optra device reporting.
15. “Longitude” – The longitude of the Optra device reporting.

Person Occupancy Plus

Easy and secure occupancy tracking and density mapping

The Foot Traffic skill provides real-time information about the total number of visitors, entering and exiting a specified location.

This information is the best indicator of the total occupancy for a given location. It can be used to identify the peak hours for customer visits, and to calculate occupancy trends across different time periods.

For the implementation of the skill, the user will specify the entry and exit regions for a particular location, via a web interface.

One of the primary sources of dissatisfaction among shoppers are delays at the checkout. 

By recording the number of customers entering and leaving the store, the staffing at checkout lanes can be scaled up and scaled down as needed. This is achieved based on real-time data and will: 
* Shorten customer wait times Improve the overall customer experience 
* Balance staffing efficiency  
* Reduce costs from overstaffing at checkout lanes. 

This skill leverages cameras and AI (Artificial Intelligence) algorithms to accurately count the number of people entering the store in real time. The cameras are placed at the entry and exit of the store to record shopper arrivals and departures.  

**How can this improve your operations?**
An accurate prediction of checkout staff level is computed through understanding the average shopping time for your store combined with an accurate count of the number of people entering and exiting the store. 

Use the information garnered from the skill to alert associates to return to the checkout lanes in advance of long queues forming, or when they are free to close their lane due to dwindling traffic. 



Adlytic.ai

Foot Traffic

**Monitor Incoming and Outgoing Foot Traffic**

HeatMap is a skill that provides a two-dimensional visualization of person traffic, color-coded based on visit frequency. 

Heatmap is the best way to visualize the utilization and visitor behavior on the floor of a store or building. It can help in identifying popular or dead areas and also help to increase sales per square feet by making data-driven business decisions.

For the implementation of the skill, the user has to specify a region, via a web interface.

Build less and do more with a library of plug and play skills

Skills Marketplace