Autonomous Security Robots – Coming to a Community Near You
A discussion with Knightscope CEO Bill Li
077 - Bill Li
At Ford Motor Company, Bill held over 12 business and technical positions, focused on 4 continents, cutting across each functional area. These positions ranged from component, systems, and vehicle engineering with Visteon, Mazda, and Lincoln; to business & product strategy on the US youth market, India, and the emerging markets in Asia-Pacific and South America; as well as the financial turnaround of Ford of Europe.
After internally securing $250 million, Bill founded and was COO of GreenLeaf, a Ford subsidiary that became the world’s 2nd largest automotive recycler. Under his leadership, GreenLeaf grew to a 600-employee operation with 20 locations and $150 million in sales. At the age of 28, Bill was the youngest senior executive at Ford worldwide.
Bill was then recruited by SOFTBANK Venture Capital to establish Model E Corporation as its President and CEO, a new car company where the “Subscribe and Drive” philosophy was first pioneered in California. He subsequently co-founded Build-To-Order Inc. (BTO) as its President and CEO, a new car company based on the direct distribution of build-to-order products.
Talking Points
- Physical Deterrence
- Collects Terabytes of Data: License Plates, Facial Detection, Cell Phone detection, Thermal Scans
- Broadcast Audio Warnings, can be Used as a Loudspeaker by Police
- Eyes, Ears and Tools for Police and Guards to do Their Job More Effectively
Connect with Bill Li
Website
https://www.knightscope.com/
Facebook – LinkedIn
John DeBevoise:
Greetings, everyone, and welcome to another serving of Bizness Soup Talk Radio. If it’s in business, it’s Bizness Soup. I’m your host, John DeBevoise. 60–some years ago, the automotive industry coined the phrase “automation.” Well, we’ve come a long ways since then. Bill Li will be joining us from Knightscope. Knightscope is a design and automation company that uses autonomous data machines used to monitor crimes in malls, parking lots, and in neighborhoods. They’re robots, folks. They’re about five feet tall, about 400 pounds, and when you haven’t seen them, they always see you. It’s fascinating to hear how this company evolved and what Bill is doing to help secure this country and you and businesses. Find out in this serving of Bizness Soup, how security has a new vision right here at the table at Bizness Soup.
Bill, welcome to this serving of Bizness Soup.
Bill Li:Thanks for having us.
John DeBevoise:It’s been a goal of mine to find out who was behind or in that robot that I saw wandering about a mall awhile back and I’m going, “What is that thing?” I’m following around and then it turned around and started following me. I guess I was of a suspicious character. You are the CEO of Knightscope, so let’s just say that you and I met for the first time at a party, at a reception, and I asked you, “Well, Bill, what is it that you do with Knightscope?” What is your cocktail version of a speed pitch?
Bill Li:Sure, so Knightscope, we build fully autonomous security robots that are helping the 2 million law enforcement officers and security professionals better secure the places you visit, work, live, and play. It’s a unique combination of self–driving autonomous technology of robotics and artificial intelligence to give a human I’ll say almost magical powers at their fingertips to better secure a location.
John DeBevoise:When you say it’s an autonomous vehicle, it’s not a car, it’s one of these odd–looking things that is wandering about performing… it could be anything. In the case that I saw, it was security detail. It’s quite innocuous in its presentation other than, what is that? So it’s able to monitor the activities in what’s going on in a mall and be able to alert someone if there is a security situation? Or is it just the eyes and ears of somebody in a monitored position?
Bill Li:There’s two things actually going on at the same time. One is we have numerous form factors and sizes of the machines, from a stationary one to an indoor one, outdoor one. If you visit us at knightscope.com, you can check them out there. The one that’s one of the most popular is the K5 Outdoor, and one of the first things we want to do from a security standpoint, which most people don’t realize, is just providing a physical deterrence can stop a lot of negative behavior.
For example, if we put a marked police car in front of your home or your office or on the highway, criminal behavior changes or your foot is going to hit that brake immediately without even thinking about it. The K5’s five foot tall, 400 pounds. It roams around on its own autonomously without any human intervention, and then the second thing is to grab as much data from the environment as possible that no human could ever possibly digest. These machines generate over 90 terabytes of data a year and we put that into with some machine learning and artificial intelligence into a format that a human can actually utilize, so it detects certain things that might be of interest for a security guard to keep an eye on.
John DeBevoise:Putting into my perspective, it wanders about and it’s able to collect all of this information around my property. Let’s say I’m a tenant in a mall and it is going about in its autonomous fashion taking record of everything that is within its scope, whether it sees it or hears it.
Bill Li:Yeah. Our clients are certainly malls, law enforcement agencies, numerous hospitals, manufacturing facilities, logistics facilities, commercial real estate, Fortune 1000 corporations, basically anywhere indoors or outdoors you might see a security guard’s a fair opportunity for us. The machines, for example, can read 1200 license plates a minute and run it against a database that the client might have.
You can detect a mobile device within a few hundred feet of the machine and treat that mobile device as if it’s a license plate. You can detect a person, you can detect the person. You can run a thermal scan of the environment. Security guards can even speak through the machine as if it’s a mobile PA system, or it can automatically issue broadcast audio warnings. There’s a lot of opportunities for the machine to do the monotonous computationally–heavy stuff and sometimes dangerous stuff that you don’t want the humans necessarily to be doing. For example, do you want to be the security guard at a hospital right now? Probably not.
John DeBevoise:Yeah.
Bill Li:Not to be funny about it, but the robots are immune and will continue–
John DeBevoise:Sure.
Bill Li:To operate despite the pandemic.
John DeBevoise:They’re built like a five–foot–tall linebacker, so not much is getting through them.
Bill Li:Yeah. We have a little bit of this challenge of being you need to provide enough of a physical presence, but this is not a military product. You can’t scare the child and you can’t scare Grandma, so it’s got to be somewhere between like, “Okay, it’s big enough that I’ll do a double–take and maybe I’ll decide not to steal that car right now because I have no idea what that thing does,” without being imposing because we’re in public settings.
John DeBevoise:Well, and you certainly couldn’t pick it up and run off with it, either.
Bill Li:No. That hasn’t happened. I’m sure it will one day, but those folks are going to be a little bit surprised when they figured out that–
John DeBevoise:Well, and it’s going to leave it’s own digital trail.
Bill Li:Exactly, and it’s basically it’s a felony in all 50 states to be tampering with the [inaudible] and we put a few people in jail and they’ll start realizing that you’re not supposed to that. No different than you probably shouldn’t mess around with a law enforcement vehicle.
John DeBevoise:We’re talking with Bill Li from Knightscope. We’re talking about automation and robotics. The term of automation really came from the automotive industry way before I was born and I saw the first steps of automation in the assembly line when the union workers were demanding so much money at the automobile factories that the manufacturers started putting in robots to start assembling the cars.
Well, now, they’ve gotten even better and bigger and more of them. Now, you’ve taken it, put it into a small package of automation and a lot of other skill sets that no one person can have and this can be in a four–wheeled vehicle in an environment that’s in the dirt from what I understand, as well as on a flat surface around a mall and parking garage. I think that’s wonderful applications.
Bill Li:Yeah, I probably refined that a little bit. I’m an ex–Ford Motor Company executive, so I’m a little bit fluent in terms of what’s going on in Detroit, and yes, there was a point in time where there was a lot of automation thrown at the manufacturing plants and this dream of having a lights–out facility with no humans in it. Life is not black and white and you realize that humans sometimes can do some extraordinary things that it would be an almost impossibility for a machine to do, and vice versa.
What’s happened in the manufacturing plant of software and hardware and humans working together to get 60 jobs an hour out the plant door is actually the same analogous in a funny way here on the physical security side of things where Knightscope focuses, where we have the machines do the monotonous and computationally–heavy work that no human could do, but the strategic decision–making and enforcement is left for the humans. A better way to think about it is these are really smart eyes and ears and tools for the officers and guards to do their jobs much more effectively.
You can contrast that to the soldiers out in a theater of war. You’ve got 2 million–plus soldiers out there, a $700 billion budget at the Department of Defense, and a Lockheed Martin and a Northrop and a Boeing building you whatever technology you might ever desire. That unfortunately does not happen for the million security guards and the million law enforcement officers.
Your listeners might want to know that the U.S. Department of Justice and the U.S. Department of Homeland Security have no federal jurisdiction over 19,000 law enforcement agencies and 8,000 private security firms. There’s literally no one in charge. There’s no innovation process. There’s no risk capital and that’s why there’s a unique opportunity for Knightscope to disrupt this very sleepy industry. We have a little bit of a crazy long–term ambition just to see if we can make the United States of America the safest country in the world, and the way to do that is through technology working hand in hand with humans.
John DeBevoise:If I came to you and I said, “I need this type of security. I’m not happy with what I’ve got,” what type of business, since this is a business show, what kind of business if I was coming to you and asking you, “I need a Knightscope robot that can take care of the security aspects or take care of my needs,” who would I be?
Bill Li:It’s funny, a lot of folks think it’s really expensive and only big–time clients like a Samsung or a Citizens Bank, or we just landed our first federal government contract, U.S. Department of Veterans Affairs, could afford this, but the effective price is actually, depending on the machine, between $4 an hour and $11 an hour, which compares favorably to some of the guarding solutions that are out there today. We do have small businesses as clients and all you need to do is go to knightscope.com and we request a private demo. My team will spend a half hour, an hour trying to understand your wants and needs and the problems that you have, and then we’ll let you know if there might be a fit that, yes, we can help you. Or we’ll be really frank with you and like, “We can’t do what you’re asking,” or we’ve got maybe a different solution to look at.
Basically, if you’ve got a crime problem that won’t seem to go away, that’s usually one of the best opportunities for us. Second, if the solution that you have today is not effective, or if you’re looking to give your guards or your officers more tools for them to cover more ground, those are all opportunities. What’s really problematic, the entire nation is focused on public safety and reimagining it, as I mentioned, there’s 2 million humans trying to secure 328 million Americans across 50 states.
John DeBevoise:Yes.
Bill Li:The math doesn’t work. They’re running 24/7, so at any point in time there’s only 500,000 people trying to secure 300 million. That just doesn’t work, and what is needed sorely is a new level of technology that can really relieve some of the tension and provide the officers and guards more capabilities so they can do some of the stuff that’s more important for a human to do.
John DeBevoise:Do these work hand in hand with the physical law enforcement or security personnel?
Bill Li:Yes. Everyone focuses on the shiny objects, the machines themselves, as we call them autonomous security robots or ASRs, but they feed all of that data into a browser–based user interface that you just get on the internet. We give you the credentials and passwords and you’re in and you’ve got access to your machines 24/7 anywhere in the world.
You’ve got an internet connection and that provides the officers and ability, or the guards, to keep an eye on things remotely. Law enforcement agents, for example, will use what we call the KSOC, the Knightscope Security Operations Center, literally in the 9–1–1 dispatch area, so they’re able to patrol a portion of the city that they don’t have the officers to cover right now.
John DeBevoise:Are all of your units mobile? Or do you have stationary ones that even blend into the surrounding areas?
Bill Li:Yeah, we have the K1 Stationary Machine that has all of the capabilities of all of the other machines. It just happens to typically sit at ingress/egress locations where there is a lot of humans or vehicles coming in and out.
John DeBevoise:You’ve got the alarm system. If somebody walks out of, say, I’m just going to use Best Buy, an they didn’t pay for the item or they didn’t get the RFID dismantled, this unit at the door would take a permanent record of whoever walked through it at that time, a picture of them and everything?
Bill Li:Yes, and keep in mind, it’s 360–degree, high–definition eye–level information and video capture, as opposed to a lot of video systems are really good at getting the top of your head and then when you go submit it for evidence, you can’t tell what the heck’s going on or if the camera’s even working or if it’s even been monitored or upgraded in the last several years. Yes, this provides literally a complete different vantage point to help with incidents and response.
John DeBevoise:Bill, I want to take a step back to when you got involved with this. How did this company come about? In every business, there’s the idea, the people, there’s the execution, the solution and such. What was the idea behind creating these autonomous vehicles? What was the genesis behind that?
Bill Li:There’s two answers for that for me. One’s personal and one’s professional. The professional answer is as an ex–automotive executive, I believe that self–driving technology is going to turn the world completely upside down. I’m just not in agreement in the path to commercialize the technology. Back in 2013, I said it’s going to take a very long time for you to get into a self–driving autonomous vehicle without a human in it and have it drive you around. I would suggest that there’s probably no one on this call to ever have done that before and is not doing it today. It’s the year 2020 and no one’s shipped anything just yet.
Part of it is because everyone’s trying to work on the most difficult problem first as opposed to what we would try to as the crawl–walk–run approach. Now that we’ve operated more than 1 million hours out in the field across five time zones, we’re coming up on our fourth winter, we’ve got this working out one to three miles an hour, and then little by little, we’ll go three to 10, 10 to 15, et cetera, and manage our way to commercializing the technology in a much faster and sane manner. That’s kind of more of a technical interest.
The personal answer is I was born in New York City and someone hit my town on 9/11 and I’m still profoundly ticked off about it and so since then I’ve dedicated my life to better securing our country. As I mentioned earlier, the way the country’s set up, as much as I love it, is structurally flawed in helping officers and guards actually have new tools and capabilities. This is why in the year 2020 you have security guards sitting in parking lot with a number two pencil and a notepad. It is horrendous.
John DeBevoise:Yeah.
Bill Li:And absolutely beneath the dignity of this nation for people to get up on our own soil, willing to take a bullet for you and your family, and the level of capability we provide to them is laughable. I think that’s what gets me up every morning and what drives me to do slightly irrational things to make this all work.
John DeBevoise:You got into Knightscope in the walk–crawl–run mode, and there’s the new technology, or not so new now, but it’s being deployed and that’s 5G. Are your units, your autonomous vehicles, are they dependent upon 5G? Or can it just be local or the 4G?
Bill Li:We’re primarily on 4G and despite all of the hoopla on TV about 5G, it’s not permeated throughout the entire nation just yet.
John DeBevoise:Your vehicles aren’t dependent upon a technology that is evolving or wouldn’t even be available?
Bill Li:5G’s going to enable us to do some more interesting things, but it’s certainly not a linchpin to commercial the technology. We’re already been shipping product for years.
John DeBevoise:You’ve been shipping product all around the world?
Bill Li:U.S. only.
John DeBevoise:Just strictly the U.S.? Okay, so you have law enforcement that uses it, malls. Airports?
Bill Li:Airports, we’re about to deploy one in Texas Airport, a rail transit area, actually a private residence in Los Angeles. We just signed a contract for water utility. We’re up on our fifth hospital I believe, so it’s again interesting opportunities. Once you start thinking about, the possibilities are almost endless.
John DeBevoise:Well, if it’s in a hospital or multi–floor, how does it push the elevator buttons?
Bill Li:Most of the crime actually happens outside.
John DeBevoise:Okay.
Bill Li:Most people think it’s indoors. It’s not. We do go up in… Like in Colorado, we’re up and down a nine–story parking structure, so the machines can go up and down that no problem.
John DeBevoise:I find that as a person, I don’t really see them. It took me a while to understand when I saw one just what it was, and at first I didn’t know what to think of it, but it had the ability to scan everything, everybody around and take a record of it. As one of our Presidents so famously said, “There is no privacy these days.” It sounds like your little guys, or your big guys, your linebackers, are collecting all of that information everywhere we go.
Bill Li:Yes, but we’re 50–state regulatory–compliant. We’re in public areas, so the concept of privacy doesn’t apply. We’re not in someone’s conference room or the restroom or something like that or someone’s private home. You’re on public property, or if our client’s private property you’re on private property, but it’s in a public area, so basically all of the technologies that are out there today are already being utilized.
John DeBevoise:As a multi–family unit landlord, on a larger project, would this be something that would work well for me as an apartment owner with my complex?
Bill Li:Yes, absolutely. HOA kind of clients or commercial property that multiple tenants or multiple residents or even a subdivision or gated community, especially if there has been a guard. The other issues is that the jobs aren’t all that consistent, so you’ve got a hundred to 400% employee turnover rates and they can’t keep the guards in the job because it’s almost not a job or a sufficiently paid well job. What we like to do is [inaudible] the guards, have more capabilities and be paid appropriately for what a human should be doing, but if there are guards already at that apartment complex or subdivision or student housing, for that matter, it’s certainly an opportunity for the technology.
John DeBevoise:Your passion is to make the environment safer and our country safer. What are some of the things that you are doing to deploy this into areas that would perhaps safeguard us from other attacks on our person and our country?
Bill Li:What we want to do is over time put thousands and thousands of these out in the field and then get them to do a hundred times more than a human could possibly ever do. For example, the machines can detect a mobile device that I mentioned earlier, right? Unless you–
John DeBevoise:Right.
Bill Li:Say a hypothetical situation, you’re in Times Square in New York and there’s a cellphone that has not moved in the last 17 hours. No human’s going to be able to detect that. Did someone lose their phone? Or is that tied to something much more nefarious?
John DeBevoise:Yes.
Bill Li:Or can you teach the machines to listen for different kinds of sounds and localize that sound? Meaning, what direction did it come from and be able to feed that with a timestamp to a human to analyze. The more you understand from the environment, the better you’re going to be able to secure it.
John DeBevoise:Well, they have technology that triangulates gunshots, where they can send officers off to a neighborhood when a gunshot is heard. Being that this will hear and see things, would it be able to assist law enforcement or other security services if there was an explosion or a gunshot and point them in the right direction?
Bill Li:Exactly. That’s on our roadmap to have the machines be able to do exactly that.
John DeBevoise:If I’ve got more than one of these roaming around, say, a mall or a casino, are they communicating or triangulating and keeping things more secure because they are able to communicate between each other?
Bill Li:Indirectly, they all feed into the same mothership application, so one of our clients that has six machines in one location, and so all of that data is being fed into that user interface and they’ve got a bird’s eye view of everything going on and where things might get tripped. For example, they’re a big proponent of facial recognition and they upload photographs of folks that they’ve had to throw out, either card counting, drug dealing, vandalism, whatever it is, and these folks, they don’t want them on their premises, so they can easily have the machines be on the lookout for them. Similarly, they can actually do that for their VIP guests and be able to say, “Hey, our whale client from last weekend just arrived at door three. Someone from guest services get down there.”
John DeBevoise:What I’m looking at is the security aspect of it, and if I am a tenant, say I’m a triple–net tenant in a mall and such and I see these wandering about, I should feel more secure because they’re keeping track very quietly of all of the activity that’s going on. If there is something nefarious that has occurred or quite possibly could occur as a result of the communication and the data that has been gathered, it’s more comforting to me as a business owner knowing that these Knightscope robots are cruising around keeping track of everybody.
Bill Li:Yeah, you don’t have to take my word for it. If you just go to knightscope.com/crime, you can see all of the data there, including data from the law enforcement agency of all of the things that the machines have been able to do, from helping a law enforcement agency apprehend an armed gunman to another law enforcement agency helping them with a domestic abuse situation, to having a security guard apprehend a thief, stopping a fraudulent insurance claim. You’ll see there that the list goes on and on and we’ve just gotten started. We’ve proven the technology works. Now, it’s time to scale it up across the nation.
John DeBevoise:There has been a lot of talk in recent months about the defunding the police departments. Does Knightscope have any opportunities? They’re talking about moving money around. Is the type of service that Knightscope provides… One of the services that the cities that are claiming to defund, is this one of the things that they may be interested in that could help them out?
Bill Li:I don’t like the word “defund,” reimagining. The system’s broken. You cannot possibly have 19,000 agencies, 8,000 private security firms all every man, woman, and child for themselves while running different processes and everything else, and none of them with any advanced tools and capabilities less our clients. For us, yes, there’s an opportunity, but it’s the same opportunity of why we started to work on this in 2013 and we’ve been almost screaming at the top of our lungs for the last seven years that this needs to get fixed.
Now, we’re grateful and excited that the entire nation is now focused on the same problem that we’ve been touting about, that there is a problem here. We can be part of that greater solution and hopefully tone down the tensions and actually start working on solutions to the problem.
John DeBevoise:Well, I have to admire the vision that you have with Knightscope and your robots, and I was fascinated when I came across them and to learn that I could have you, the CEO of the company, on my show to talk about what they are, where to find them, and if I was a private citizen, as you said, that in Los Angeles somebody is getting one of these. What does a private citizen need with one of these? Or is it just because they can?
Bill Li:It’s someone that does actually have a genuine need. It’s not a toy for them. It’s a large property and-
John DeBevoise:Okay. Well, I’ve got large properties, too, but I’m not quite sure… They have no arms. I can’t put a rope on them and tell them to go get that steer. That’s a different version.
Bill Li:For a small fee, we can work on that.
John DeBevoise:We can get some arms on it, yes. Bill, I want to thank you for being on this serving of Bizness Soup. It’s been a pleasure and it’s an honor to have you come on and talk about this amazing technology that you have implemented into Knightscope. Bill, thanks for being on Bizness Soup.
Bill Li:Absolutely. Thanks for having us and don’t forget to visit knightscope.com and please be safe out there.
THANK YOU for visiting BIZSOUP