Moorebot Scout-E: A detailed look at features and functions.

Image: Moorebot Scout Model E

Introduction

TheMoorebot Scout model E is a small, AI-powered robot that can be used for intelligent monitoring of your home or office. It features a 1080P camera with night vision, 2-way audio, and a variety of sensors that allow it to navigate around your space and detect people, pets, and objects.

Some of the features that make the Moorebot an exciting prospect for home survellance are detailed below.

Camera: The Scout model E's camera has a resolution of 1080P and can capture clear detailed images and videos, even in low light conditions. The camera also has a 120-degree field of view, so you can see a wide area.

Audio: The Scout model E has a built-in microphone and speaker, so you can communicate with people or pets in the room.

AI features: The Scout model E uses AI to detect people, pets, and objects. This allows it to automatically patrol your space and send you alerts if it detects anything unusual.

Automatic patrol: The Scout model E can be programmed to automatically patrol an area.

Voice control: The Scout model E can be controlled using voice commands through Amazon Alexa.

Scratch programming support: The Scout model E supports Scratch programming, which is a visual programming language that is easy to learn.

Wi-Fi connectivity: The Scout model E connects to your home Wi-Fi network, so you can control it from anywhere when paired with the Moorebot cloud service.

Battery life: The Scout model E has a battery life of up to 2 hours allowing it to go for long periods of time without recharging.

The Moorebot Scout model E is a versatile and powerful robot that can be used for a variety of purposes, including:

• Home security
• Pet monitoring
• Eldercare
• Business security
• Education
• STEM learning

The Moorebot Scout-E is an exciting potential product, but my initial testing found that it did not meet expectations. I decided to conduct more thorough testing to explore all of its features and see if it would surprise me. The Scout-E is advertised as an AI-assisted security camera on tracks, so I wanted to determine if I was not giving the Scout-E enough time to learn from its mistakes and become a more usable platform. There are many companies that use the term "AI" to sell their products even though they do not actually utilize AI in any form

My testing of the Scout-E platform aims to determine whether it is truly AI-assisted that can learn to avoid obstacles and accurately calculate its way back to the charging station. The documentation does not specify the type of AI used, whether it can learn, or how it detects obstacles. My testing should help shed light on how the Scout-E operates.

Testing Methodology

The testing methodology will be real world testing where I will go through each advertised feature and test it multiple times. We will start with the patrol path feature testing a simple path, moving on to a more complex path. I will start with path creation followed by three runs of these paths to see if the Scout-E can complete the patrol successfully. During testing I will be observing the time it takes to complete the run and how many times the Scout-E gets off course bumping into obstacles that it should be able to avoid. Each of the paths will be in an open course with only minimal obstacles.

Once I complete the patrol path tests I will move on to testing its ability to follow a human and an animal. I will run three tests with the Scout-E following both humans and animals. During the testing I will test with several subjects in the room and a test with just a single subject to see if it makes a difference.

Testing Results

Creating a Patrol path

First attempt

My first attempt at creating a patrol path didn’t go so well, the controls for Scout-E were challenging. There were times where the Scout-E would over steer causing it to go off the path I was trying to create. Once I was able to get familiar with the controls I ran into another issue where the Scout-E wasn't able to dock with the charging station from a distance of more than a foot away. Instead of docking the Scout-E turned 180 degrees and started to move away from the station. I eventually took manual control and guide it back to within less than a foot of the station where it was able to dock and charge.

Second attempt

On my second attempt I was able to successfully create a patrol path and save it once I drove it back to the charging station. In the documentation it talks about being able to drive the Scout-E to the half way point of the patrol path once you get to your destination you can hit the return button sending the Scout-E back to the charging station. In the testing I was not able to have it automatically return to the charging base, it attempted to head back in the direction it started from but never made it to the charging base. Each time I wanted to create a patrol path I had to drive the Scout-E around the entire path returning it to the charging base so that it could dock.

Running the patrol path

Image: Scout-E bumps into object

First attempt

The Scout-E struggled to run the patrol path, first getting stuck in the starting room, it took some time to figure out how to get out the door. Once the Scout-E was outside the room it ran into a wall while following the predefined path. I let the Scout-E run for a while but it never actually made it to its destination. The bot struggled, getting turned around and running into walls on multiple occasions.

Second attempt

On the second attempt to run the simple patrol path the Scout-E was able to make it to the half way point but once it made it back to the room where the charging station was located it ran off path getting stuck. I let the Scout-E try to work out where the charger was but eventually it gave up and had to be manually driven back to the charging station so to self dock.

Third attempt

My hope on the third attempt was that the Scout-E would learn from the first two attempts but this attempt went even worse then the first two. As soon as the Scout-E left the charging station and exited the room it went completely off course ending up going in the wrong direction all together. I again allowed the Scout-E attempt to see if it could figure out how to get back to the charging station but eventually I found it stopped against a wall.

After the third attempt to run the simple patrol path I decided to move the charging base to another location that was a wide open area. Once I moved the Moorebot Scout-E to another location, I ran my tests again a couple of times. The first test went really well with the Scout-E finishing the patrol.

Complex Patrol Path

Image: Scout-E screenshot

First Attempt

Since the Scout-E was able to achieve success from the new location on two seperate occasions I felt a more complex patrol path would better test its abilities. From the same location I created a path that went on a complete loop around the kitchen and returned home to the charging base. The Scout-E started out in a good direction but once it made the first turn into the kitchen things went wrong. The Scout-E strayed from the path and ended up off track, once it was off the path the issues worsened until the Scout-E got stuck in a corner. I let the Scout-E try to figure out how to get back from its location and after many attempts the Scout-E wasn't able to figure out how to get back to its charging station.

Second Attempt

On the second attempt to follow the complex patrol path the Scout-E steered off path and bumped into a few things but eventually recovered making it back to its charging base.

Third attempt

On the third attempt the Scout-E again ended up off course but this time it got stuck in a corner. After being stuck in the corner for a few minutes the Scout-E was able to start heading back towards the charging station, unfortunately it missed a big chunk of the path it was supposed to follow. Once it finally made some progress towards returning to the charging station the Scout-E ended up confused and stuck. The Scout-E spent several minutes trying to figure out how to get back on path. While watching the Scout-E try to figure out where to go it looked alot like the Scout-E was stuck in a loop making the same movements over and over without success. I eventually took over and manually drove the bot back to the charging station where it was able to self dock.

Tracking people and animals

Image: Scout-E tracking dog

First attempt

The first attempt at tracking a human was somewhat successful depending on how quickly the person moved. The Scout-E lost track of the human on several occasions. Following a human around corners even at a slow pace seemed difficult for the Scout-E. There were several times that the Scout-E lost track of the subject and veered off course into a wall or corner. When tracking a person or animal make sure there is only one subject in view of the Scout-E. The Scout-E is easily confused when there are multiple subjects in its limited view. I tried to switch over to a dog to see if this task would be harder and the Scout-E was even less successful. The Scout-E was able to track a dog when it wasn’t really in motion but as soon as the dog moved it would lose track of it quickly and started to scan the area for motion.

Second attempt

The second attempt at following a human went much better this time the human was alone without any distractions like animals that could catch Scout-E's attention. As long as the human walked at a slow pace the Scout-E was able to keep up only losing track of the subject when going around corners. The Scout-E was usually able to pick back up when the subject would stop or back up a little but for the most part this test was a success.

Third attempt

The third attempt was a bit of a mixed bag, I had the Scout-E follow a child, again for the most part as long as the subject was moving at a slower pace it was able to keep track and follow along. The Scout-E ran into a few issues when the subject went around a some corners and even bumped into furniture at one point. I would have expected the Scout-E to detect the furniture before running into it but it didn’t.

Conclusion

Test Findings

The Scout-E struggled to run the patrol path in the first three attempts. It got stuck in the initial room, ran into walls, and went off course.

When the bot’s charging station was moved to a new location, it was able to successfully run the simple patrol path twice.

However, when the bot was given a more complex path, it got off track or stuck in a corner. The Scout-E was not able to figure out how to get back to its charging station once it was off the patrol path.

The Scout-E also had difficulty tracking people and animals. It lost track of humans when they moved too quickly or went around corners. It was also easily confused when there were multiple subjects in its view.

However, the Scout-E was able to track a human successfully when the human was walking at a slower pace and there were no distractions.

Overall Impression of the tests

The Scout-E has some challenges with navigation and obstacle avoidance. It is not yet able to reliably follow a patrol path or track people and animals in complex environments. However, it shows promise and could be improved with further development.

Here are some specific recommendations for improvement:

1. The Scout-E's navigation algorithm could be improved to better handle complex paths and obstacles.
2. The Scout-E's sensors or AI could be improved to better detect objects in its surroundings.
3. The Scout-E's tracking algorithm could be improved to better track moving objects specifically when going around corners or when multiple objects are around.

With further development, the Scout-E could be a valuable tool for a variety of applications, such as security and surveillance but as it stands today it needs work to be usable.

Final Thoughts

The Moorebot Scout-E is advertised as an AI-assisted camera on tracks with several features that make it an attractive platform for surveillance and security. However, my testing found that the Scout-E falls short of its promised capabilities. The AI capabilities are rudimentary, and the robot's ability to track and follow people or animals is not as advertised. Additionally, the Scout-E had difficulty following a predefined path while avoiding obstacles and recalculating its path back to the charging station. I had hoped that the AI assistance would learn over time and improve with repeated use, but this does not seem to be the case. Overall, I am disappointed with the Scout-E's performance and do not recommend it for use as a surveillance or security robot.

If you are looking for an out-of-the-box solution that performs extremely well, the Moorebot Scout model E is not the right solution for you. However, the open-source nature of the Scout-E provides a potential opportunity for improvement. With some work and skill, the robot's capabilities can be enhanced. I plan to explore this possibility in an upcoming blog post, where I will detail my findings and how I was able to improve the Scout-E.