Many people, including the public, the media, investors, suppliers, and even prospective clients, are unknowingly and inadvertently blinded by seeing autonomous security robots in action and, as a result, do not immediately consider the wealth of data being collected. The robots tend to get all the glory because they are the big, shiny object, but the real beauty to end users lies in the intelligence that they gather, which must be delivered in a thoughtful and quickly digestible manner. The tool used to do that is a custom-made user interface that we created from the ground up known as the Knightscope Security Operations Center, or KSOC.
The KSOC is included as part of the Machine-as-a-Service (MaaS) subscription we offer to clients across the country. It aggregates data from the robots, analyzes it for anything out of the ordinary and, if needed, alerts humans so they can take action. It is the conduit for maintaining communication between the robots and our customers and is a key component of our “SOFTWARE + HARDWARE + HUMANS” strategy.
Today’s KSOC utilizes artificial intelligence, sophisticated databases and protocols, and even leverages cloud computing, but it was not always that way. So here is the (slightly painful and sometimes cringe worthy) story of how the KSOC came to be.
When the company first started in 2013, we asked ourselves how our users might be able to control the basic functionality of the technology and see what is happening in real time through the eyes of the robot at any location. Obviously, the available video management systems would not work because they were essentially only able to ingest information. In order to fully serve our clients’ needs, we needed to build a two-way software platform from scratch to work in conjunction with the robots. Thus, the KSOC was born. For fun, we even dreamed of building an actual KSOC module for security professionals to sit in!
We studied other security tools and immediately knew ours would need to be better. We had many unique functionality requirements that were going to take a tremendous amount of time and resources to resolve, and we still needed to run the company and fundraise at the same time. So, we did what any scrappy entrepreneur would do – we created a PowerPoint presentation (actually Apple Keynote) in which we embedded several videos to illustrate to potential customers and investors how the KSOC would be used in the future. It’s a little embarrassing, but here’s a snapshot of our ‘genius’ idea setup at the Plug and Play Tech Center (accelerator we were in at the time) as well as on stage at the LAUNCH Festival in front of thousands of people on a massive screen (insert face palm emoji here).
Even as early as it was (and knowing full well that the technological capabilities were not there yet), we had videos of people with red boxes around their faces because we knew that we would someday want to offer facial recognition as an optional feature (which we now have available on our K1 Stationary machine in beta). It was actually a very cool demo if you ask us… and a LOT of the folks we shared it with agreed!
The first function we needed to accommodate was streaming video from the robots’ cameras to a local browser. We initially used a Windows Explorer version of a single page HTML file that used the snippet of code that the camera vendor provided. It might not seem like much now, but it was a tremendous victory for us at the time and we were very excited with KSOC version 1.0!
We elected to take things a step further for versions 2.0 and 3.0 by making an application based off of a well known web framework. We tapped in our lead designer with the following request: “It needs to be beautiful, navigate seamlessly and provide a ton of information.” Not an easy feat, but he was up to the task and got it done. We were now on our way to the next level.
Everyone that we showed the application to provided the same positive feedback, “it is beautiful!” Some told us that it was “a bit complicated,” and that there was a “lot of information on one screen.” This was all great input that we took to heart. We also received numerous requests for mobility, so we began to build mobile apps. Showing potential customers and investors the KSOC on mobile devices was the part of the demo many of us loved the most. They were always in awe – and sometimes disbelief which was so cool to witness.
So, we added more and more functionality like Machine Learning (ML) and Natural Language Processing (NLP) to versions 4.0 and 5.0 of the application, and did more work on things like scalability and redundancy as our client list continued to grow.
Then we decided to take all the client feedback we had received up to that point and create what is now KSOC v6.0!
We did things very differently for v6.0 – we assembled a small, special team with representatives from each functional area of Knightscope and applied the principles outlined in the book “Sprint: How to Solve Big Problems and Test New Ideas in Just Five Days” by Jake Knapp. We followed it to the T, and in one week we were ready to show our designs to a set of existing clients for feedback.
The team was worried. What if our clients do not like it? What if they loved it? Should we prepare our evaluators before we show them the screenshots and working mockup? To our elation, they loved our designs! They used terms like “very easy to navigate,” “very intuitive,” “easy to use,” etc. We decided right then and there that this was the tool we would build next.
The fact that we have been in business for a while and that we always incorporate client input into what we do has made us successful at building tools that meet client needs – and since we operate 24/7/365 we get A LOT of feedback – the good, the bad, and the ugly.
We will continue learning and iterating and bringing tools to life that not only help our clients but delight them. Technology changes so quickly that we’ve already started to dream about what version 7.0 of KSOC might be!