Wearable Cameras will be Improved with the Addition of AI Technology

After Motorola Solutions invested in artificial intelligence (AI) company Neurala in January 2017, the two companies have announced plans to bring AI to devices such as wearable cameras. Neurala’s AI software, called “working at the edge”, allows existing devices, such as Motorola’s Si500 body-worn camera, to have AI functionality, allowing the devices to search for certain objects or people. The software can adapt an existing device’s hardware to allow them to leverage AI functionalities and continuously learn the more they are used. 

The two companies are focusing on adding this technology to wearable cameras used by law enforcement agencies, to prevent officers from manually searching for objects or persons of interest, either in the field, or from the hundreds of hours of recordings from wearable cameras. The network can push a photo of, for example, a missing child to a network of thousands of wearable and dashboard cameras used by the police, which then uses the AI technology to search for the child within all the camera’s recordings. If a match is found, an alert is sent to ensure that the nearest officer can find the child. This operation is done without the need for human interaction, allowing it to be automated and preventing the need for officers to look at every face in a crowd or do repetitive work such as searching through the hours of recordings. This not only improves efficiency, but also reduces the chance of false positives and other human errors.

The technology has several other applications within and outside of the law enforcement industry. The AI technology can also be leveraged to search for suspects, as well as certain objects, such as backpacks or vehicles. The technology can also be added to other, non-wearable camera devices, such as security cameras, dashboard cameras, and drones. For example, Neurala has been working with the Lindbergh Foundation to help fight elephant and rhino poaching in southern Africa, adding its AI technology to drones as part of its Air Shepherd program. It searches through the video, even in infrared, in real-time to detect animals, vehicles, and poachers around the clock.

Neurala’s and Motorola’s partnership is helping to make devices evolve from a passive sensor to an intelligent device; however, the technology still requires improvement and is not 100% accurate. That said, AI does have the ability to approximate up to 90% of a human’s capability to recognize an object, with the added advantage of being able to function 24/7.

Motorola and Neurala aren’t the only companies looking to bring this AI functionality to wearable cameras. In February 2017, wearable camera company Axon acquired two AI teams with a focus on video footage – startup Dextro and Fossil Group’s computer vision division. Axon is planning to leverage the knowledge and technology from the two companies to bring AI technology to its wearable cameras. The technology will have the ability to extract information from the video and automatically transcribe speech, which will allow an officer to check over what a suspect said in the field. The videos will also be intelligently redacted, blurring the faces of bystanders to allow the footage to be publicly released. The AI technology will also allow officers to search for a specific item or person, such as a “red Ford” or “a white male in a green jacket”.

Wearable cameras already have a range of benefits, especially in a law enforcement environment, providing a source of evidence against suspects, as well as ensuring officers follow best practice. Adding AI functionalities will help to make the devices even more advanced, saving vital workforce hours when it comes to searching for a person or object, blurring out innocent bystanders, and transcribing what has been said. With many companies looking to add the technology to the devices, shipment numbers are expected to see a large increase, with nearly 24 million of the devices being expected to be shipped to enterprise in 2022.