Trending Topics

What TASER’s acquisition of 2 AI companies means for the future of policing

CEO Rick Smith said that he envisions the company building tools for Axon customers that will leverage the machine learning and artificial intelligence capabilities of the new team of experts

axon.jpg

The team of roughly 20 researchers and engineers will accelerate the introduction of new AI-powered capabilities for public safety.

TASER International acquired two companies that specialize in machine learning — Dextro, Inc. and Misfit, Inc. — which will be combined and leveraged to form a new artificial intelligence group called “Axon AI” within TASER International. The team of roughly 20 researchers and engineers will accelerate the introduction of new AI-powered capabilities for public safety. Financial terms of the transactions were not disclosed.

“There is massive interest in AI right now, from all major tech companies,” said TASER International CEO Rick Smith in an exclusive interview with Police One. “Whether it’s Elon Musk or Bill Gates, many of the tech thought leaders are pointing out that everything that we’ve been building in software up until now has been leading to this. Self-driving cars are becoming a practical reality, and you couldn’t do that without AI. Google rebuilt their entire Google Translate in nine months using AI, and its performance significantly improved over the old version — which was built by hundreds of engineers.”

When you take a step back and look at AI, it is really just a different way of developing software. Whereas in the past you would have teams of human developers coding very specific instructions — ‘if this, then that’ — with AI, the software itself actually learns and codes itself through learning and observation.

Smith said that he envisions the company building tools for Axon customers that will leverage the machine learning and AI capabilities. For example, as officers are manually going about the process of redaction, there may be a time in the future where the software actually observes and learns from what the officer is doing manually, so that the next day, the automated redaction capability is actually better, faster and more accurate.

“We’ve got all these videos, and we’ve got a hundred-thousand-plus officers that are working and analyzing these videos,” Smith said. “That feedback loop can become incredibly powerful.”

Changing the workflows

Smith said that the goal is that every time a police officer interacts with digital information, the law enforcement community as a whole gains value from that interaction. Smith said that police spend a lot of time on data entry and not enough time culling data from what’s potentially available to them. He and his team at TASER believe that machine learning and AI can help solve that problem.

“Most of the time police officers are interacting with technology, they’re putting data in. We believe that this can be largely automated. We’re already recording these incidents. All the information needed for a police report is in the videos. So we see the opportunity with AI to automate all the things cops have to do to process that information manually. One of those things is redaction — that’s probably one of our first, easiest wins. But another is transcribing the video into a records management system — creating an RMS system that’s really built around video. Our mission is that police records in the future should be recorded, not typed.”

With AI, the workflow fundamentally changes. The wearable device records the event. The data is sent to the cloud where it is categorized, labeled and stored. The video is transcribed into an RMS and made available to approved users. With potentially millions of terabytes of information being recorded and stored in TASER’s cloud-based Axon Network solution being processed in an automated fashion (as opposed to manually), it becomes possible to envision a world where police are given countless more hours to be out on the street — not writing reports.

“We just need tools to extract the information from the recording, and that’s where AI comes in,” Smith said. “These are perfect applications for machine learning over the next several years, and we felt we’re in a unique position these two teams felt the same way that this is a really interesting problem to get to work on. I mean, what’s more interesting than police work? And if you’re going to do video analytics, what could possibly be more interesting that police videos?”

The cloud is the key

TASER recently released a report examining technologies that will shape the future of law enforcement and not coincidentally, one of the areas of focus was the nexus between AI and cloud computing.

The report said that one of the key advantages to the cloud is having access to so much data processing power. Using artificial intelligence and hyper-scale computing agencies can rapidly process digital video, extracting and logging important information automatically.

In order for AI to really work well, there needs to be two key elements: a lot of data, and a lot of people using that data in a way that the software can learn from. This is where TASER fits right in.

“What we’re doing in the Axon business is all about building a network of devices, data, software and people. We’ve got all of this law enforcement information with these videos, which is one of the richest treasure troves you could imagine for machine learning,” Smith said.

With the combination of cloud computing and AI, law enforcement can now ingest data from several thousand cameras and automatically detect and flag anomalies as they occur.

Looking into the future

Smith said that in the future, AI and machine learning can also tap into closed-circuit TV security cameras and other sources of video to help find fugitives and missing persons. An officer could conceivably enter into the system a search for a specific vehicle with a driver who fits a certain description and have the system search hundreds of thousands of hours of recorded and real-time video for that specific person of interest (as opposed to having an officer sit there and do it).

Other potential applications for AI would be in predictive policing. By allowing a machine to observe and learn from various behaviors and incidents recorded on video, police leaders could be provided more accurate information to inform how and where and when they deploy officers to prevent — and not just respond to — crime.

“Imagine having one person in your agency who would watch every single one of your videos — and remember everything they saw — and then be able to process that and give you the insight into what crimes you could solve, what problems you could deal with. Now, that’s obviously a little further out, but based on what we’re seeing in the artificial intelligence space, that could be within five to seven years to start seeing that level of intelligence in these systems,” Smith said.

It is clear that these acquisitions help to will broaden TASER’s Digital Evidence Management System to enable law enforcement agencies to gain more insight from the massive volumes of digital evidence they already have and will continue to gather for many years to come.

Doug Wyllie writes police training content on a wide range of topics and trends affecting the law enforcement community. Doug was a co-founder of the Policing Matters podcast and a longtime co-host of the program.

RECOMMENDED FOR YOU