Qualcomm Technologies, Inc. first partnered with Axon in 2017 when the company was searching for a new imaging platform for a connected body-worn camera to enable real-time situational awareness. This led to the introduction of the Axon Body 3 with Respond map updates and livestreaming video based on the 8053 system-on-chip (SoC). We helped Axon build a foundational platform for camera development at that time as well.

AI for a new era

Axon conducted extensive technology and partner research before architecting its Fleet 3 in-car video solution, with all roads pointing to a new design. When defining Fleet 3, Axon envisioned a distributed and expandable architecture that would provide flexibility on deployment configurations and be suitable for automotive installation and use.

It needed powerful SoCs to perform various functions including one capable of image processing and edge artificial intelligence (AI) to introduce an Automatic License Plate Recognition (ALPR) solution. This innovative integration for ALPR was to be developed from the ground up for best performance and would provide actionable data to officers in the field.

The team’s ambitious specifications were vision-centered on the most advanced dash camera possible, including:

• Building a new camera system architecture
• Enable real-time livestreaming video similar to Axon Body 3
• Adding ALPR capability
• Replacing Wi-Fi camera connection with Ethernet
• Eliminating dependency of personal computer to upload videos

The real-time capabilities and license plate scanning were particularly ambitious problems to solve having never been done in this way.

As a result, Axon initiated exhaustive due diligence to identify chipsets aligned with its desired specs, with an emphasis on key strategic business and product considerations, especially:

• Device AI processing
• Seamless image processing
• Efficient data transfer
• Porting software reuse
• Cost-effectiveness
• Roadmap support

After the successful build of Body 3, Axon had Qualcomm Technologies in sight for the new Fleet 3 architecture — though the decision was not straightforward.

“When that investigation started, we spent several months analyzing the pros and cons of chipsets available in the market — everything from features, performance, cost, roadmap, and development support,” recalls Blake Bullock, Senior Director and General Manager, Fleet Products for Axon.

The team narrowed down the options on the imaging chip to two SoCs in the vision intelligence platform for the cameras, but there was a big unknown in what to use for the core digital video recorder (DVR) processor.  The project gained clarity and focus when the group sat down with us for a bespoke presentation at CES.

“When Qualcomm shared what its Internet processor family could do, my colleague and I looked at each other and knew right away that Qualcomm was the right fit,” he says.

Vetting began, and within three weeks, Axon selected the Qualcomm Vision Intelligence Platform and Qualcomm Internet Processor (IPQ) platform for the cameras and DVR respectively.

End-to-end collaboration

Foundation in place, Qualcomm Technologies and Axon worked together in a proficient manner, hosting in-person workshops with the electrical engineering, software, and design teams over the span of one year, working through questions and requests along the way. Axon’s use cases required strong collaboration in working through challenges, most notably due to new product features and the need to ensure seamless video frames even with AI continuously scanning high-resolution stills for license plates, and the camera sending multiple encoded video streams for recordings and livestreams.

“Qualcomm went beyond the normal support,” Bullock adds. “The engagement was great; we have nothing but positive things to say about how the team provided us with timely answers at key stages of the product to keep us moving along, sometimes answering questions we didn’t even know to ask yet.”

AI edge processing for more response, performance, and sales

Qualcomm Technologies enabled Axon to achieve the end-product performance it envisioned, driving it to surpass go-to-market goals.

One way edge AI processing changed the game is seen in the new ALPR technology in the Fleet 3 in-car camera system. This feature gives law enforcement real-time alerts of license plates associated with crimes and missing persons. The system is built-in to the in-car video system which reduces complexity of the equipment to install and reduces the total cost of ownership.

Fleet 3 sets a new bar for in-car video performance with advanced in-car video system features, an AI-powered, fully connected ALPR service, and real-time situational awareness technology. As Axon’s first proactive crime prevention tool, AI takes the company’s mission to the next level.

Looking forward, both enterprises are excited to collaborate again to develop an even safer future.

“We appreciate the partnership with Qualcomm to find appropriate processor solutions, and then helping us to achieve our product requirements and goals,” he says, concluding, “We look forward to collaborating on future innovative projects.”