1 00:00:04,080 --> 00:00:08,919 Everybody has the same problems, and that's the distribution of trust. How do 2 00:00:08,960 --> 00:00:14,320 you trust unmanned devices millions of amount in operation and how do you coordinate them? 3 00:00:14,320 --> 00:00:27,000 How do you get them all working together? Welcome, everyone's the Industrial 4 00:00:27,039 --> 00:00:31,440 Security Podcast. My name is Nate Nelson. I'm here with Andrew Ginter, 5 00:00:31,679 --> 00:00:37,079 the vice president of Industrial Security at Waterfall Security Solutions, who's going to introduce 6 00:00:37,119 --> 00:00:40,200 the subject and guest of our show today. Andrew, how are you. 7 00:00:41,079 --> 00:00:44,399 I'm very well, Thank you, Nate. Our guest today is doctor Christopher 8 00:00:44,479 --> 00:00:48,880 Gorrog. He is the CEO at block Frame, Inc. And he himself 9 00:00:48,960 --> 00:00:53,880 is the host of the new cyber Frontier podcast and he's talking to us today 10 00:00:54,039 --> 00:01:00,840 about establishing trust distributing security credentials in large industrial arts operations. Then, without 11 00:01:00,880 --> 00:01:07,680 further ado, here's your conversation with doctor Gorog. Hello Chris, and welcome 12 00:01:07,719 --> 00:01:11,480 to the podcast. Before we get started, can I ask you maybe to 13 00:01:12,000 --> 00:01:17,239 give a few words of introduction about yourself and about the good work that you're 14 00:01:17,239 --> 00:01:21,439 doing at block Frame. Yeah, thanks a lot, Andrew. My name 15 00:01:21,519 --> 00:01:26,879 is Christopher Gorrog. I am a PhD in cybersecurity, just recently finished my 16 00:01:26,000 --> 00:01:33,319 degree where I worked on the distribution of trust by using all means of other 17 00:01:33,400 --> 00:01:38,719 things blockchain, what we call distributed ledger. But my background has been in 18 00:01:38,760 --> 00:01:44,359 cybersecurity. I host the new cyber Frontier podcast, which we have over four 19 00:01:44,400 --> 00:01:48,239 hundred episodes, been running nine years, probably one of the largest followed out 20 00:01:48,239 --> 00:01:52,519 there. I chaired Digital Privacy for I Trip we for a global initiative, 21 00:01:53,280 --> 00:01:59,400 among many things, hold several patents and my work with the University of Colorado 22 00:01:59,560 --> 00:02:05,959 and this State of Colorado to help support legislation in the area of privacy and 23 00:02:06,000 --> 00:02:09,000 security. So a lot of things we'll get into as we unpack in today's 24 00:02:09,400 --> 00:02:15,319 Today Show. You know, our topic today is trust distributing security credentials in 25 00:02:15,840 --> 00:02:20,199 large industrial operations. Can you talk to me a bit about the problem. 26 00:02:20,199 --> 00:02:23,560 What is a large industrial operation? What does it mean to distribute security credentials? 27 00:02:23,599 --> 00:02:28,520 Why is this important? Yeah, Andrews, So, I mean, 28 00:02:28,599 --> 00:02:34,199 that's that's the ultimate question. The problem set that I've been solving or working 29 00:02:34,240 --> 00:02:38,879 on literally my entire career, because even in early stages of my careers working 30 00:02:38,879 --> 00:02:44,360 with many different vendors that are implementing security. Everybody does it a little bit 31 00:02:44,360 --> 00:02:50,199 different. Everybody has the same problems and that's the distribution of trust. How 32 00:02:50,199 --> 00:02:54,439 do you trust unmanned devices millions of them out out in operation, and how 33 00:02:54,439 --> 00:02:58,360 do you coordinate them, how do you get them all working together? And 34 00:02:58,400 --> 00:03:04,039 then how do you secure Fear of them becomes a function of that trust once 35 00:03:04,120 --> 00:03:08,759 you can distribute it. We have several methods that of distribution of trust using 36 00:03:09,360 --> 00:03:14,719 certificates and PKI, which the parts that we have a problem with is the 37 00:03:14,800 --> 00:03:19,639 human part. The people that own the trust authorities, the certificate authorities, 38 00:03:19,879 --> 00:03:23,879 the people that we have to rely on, and they don't trust each other, 39 00:03:23,039 --> 00:03:27,960 they don't work together. They're different governments, they're different companies, and 40 00:03:28,000 --> 00:03:35,800 we have a disjointed So looking at solving the problem industrywide is what I've been 41 00:03:35,840 --> 00:03:38,400 approaching literally for you know, some of my work started in two thousand and 42 00:03:38,400 --> 00:03:43,759 six, two thousand and seven working with SGIP and the s martkeret in our 43 00:03:43,800 --> 00:03:49,759 optability panel. You'll find me on the Cryptographic Key Management of the Nister seventy 44 00:03:49,800 --> 00:03:53,759 six twenty eight IR as one of the initial authors where we put together you 45 00:03:53,800 --> 00:03:59,719 know, what is needed for cryptographic key management. Back then, we didn't 46 00:03:59,719 --> 00:04:04,199 have of the global recognition of the tools that we know in blockchain. So 47 00:04:04,639 --> 00:04:10,719 my work in my dissertation has been in trying to solve the problem of trust 48 00:04:10,719 --> 00:04:15,920 distribution by taking care of the human needed pieces using distributed ledger and making that 49 00:04:16,000 --> 00:04:20,519 work. And that's what the focus has been. And I think we have 50 00:04:20,600 --> 00:04:28,519 some great solutions where we can show now feasible, global, attainable, scalable 51 00:04:29,199 --> 00:04:32,879 problem solutions in that area. And you know, to clarify, you've said 52 00:04:33,000 --> 00:04:38,399 millions of devices. I mean I understand cell phones, I understand smart watches, 53 00:04:38,439 --> 00:04:41,800 millions of these things in the industrial space. You know, to me 54 00:04:41,879 --> 00:04:45,959 when we're talking millions of devices, what leaps into mind is the smart grid. 55 00:04:46,079 --> 00:04:49,720 It is you know, smart meters, it is advanced metering infrastructure. 56 00:04:49,759 --> 00:04:53,439 Whatever the buzzword of the day is. Is that what we're talking about here 57 00:04:53,519 --> 00:04:59,519 or other other applications. Yeah. Absolutely, We're talking from the smallest sensors 58 00:04:59,800 --> 00:05:03,519 to even PCs and servers, so all across the board. And that's one 59 00:05:03,560 --> 00:05:13,000 thing. For years I worked with a semiconductor manufacturer in placing and putting cryptology 60 00:05:13,399 --> 00:05:18,360 into many different vendors devices. And you know, in that realm there's literally 61 00:05:18,399 --> 00:05:24,199 like one hundred you know, sixteen bit processors made for every thirty two bit 62 00:05:24,240 --> 00:05:29,279 processor, and there's like one thousand and eight bit processors made for every every 63 00:05:29,319 --> 00:05:31,959 sixteen bit processor. So the smaller the devices than actually the more of them 64 00:05:32,000 --> 00:05:38,600 there are out there, and that's the problem. We have relatively efficient solutions 65 00:05:38,879 --> 00:05:44,399 for the computers that humans sit behind, but the millions and billions of devices 66 00:05:44,480 --> 00:05:47,240 that are unmanned on the grid that communicate as a mesh that we might just 67 00:05:47,399 --> 00:05:54,399 read on off temperature power from one hundred times a day or even an hour, 68 00:05:55,279 --> 00:05:59,920 that that's really where the problem is is in that mass amount of devices 69 00:06:00,040 --> 00:06:04,759 out there. So Nate, you know, I'm thinking back. I remember 70 00:06:04,879 --> 00:06:10,000 a few episodes ago there was a gentleman on talking about the can bus. 71 00:06:10,000 --> 00:06:13,319 I'm not sure our remember his name, but that's the bus that's in automobiles. 72 00:06:13,399 --> 00:06:17,439 I think this was Ken Tindall, Yes, And you know his observation 73 00:06:17,519 --> 00:06:23,240 of truism in the space is that encryption is a tool for turning every problem 74 00:06:23,240 --> 00:06:28,759 into a key management problem. And I think this is sort of another example 75 00:06:28,800 --> 00:06:33,040 of that that principle that we're seeing here. You know, we're talking about 76 00:06:33,120 --> 00:06:39,600 blockchain, we're talking about cryptography, we're talking about authentication, so that devices 77 00:06:39,680 --> 00:06:45,040 can prove they are who they are, and all of that involves managing keys, 78 00:06:45,240 --> 00:06:50,279 it involves more generally managing trust and the application that springs to mind. 79 00:06:50,319 --> 00:06:54,160 The first place I saw this in the industrial space, you know, a 80 00:06:54,279 --> 00:06:59,720 decade ago, was in the context of smart meters, power meters on every 81 00:07:00,040 --> 00:07:05,720 every consumer's home or apartment, you know, billing appliances, and we're talking 82 00:07:06,000 --> 00:07:10,839 big systems. We're talking you know, a distribution power distribution system in a 83 00:07:10,879 --> 00:07:15,120 city with you know, three million smart meters, and you know, as 84 00:07:15,199 --> 00:07:20,279 Christopher pointed out, there's other really big systems out there. So you know 85 00:07:20,360 --> 00:07:26,199 what springs to mind is if you're monitoring, let's say, a water treatment 86 00:07:26,240 --> 00:07:29,680 system and you want to know what's coming at you in the watershed, you 87 00:07:29,759 --> 00:07:35,360 might have rain fall sensors. You might have you know, water level sensors 88 00:07:35,399 --> 00:07:42,800 spread in thousands of locations throughout a massive, you know, three hundred five 89 00:07:42,879 --> 00:07:50,079 hundred kilometer wide and long watershed. You might have weather sensors throughout you know, 90 00:07:50,160 --> 00:07:55,920 a whole country, gathering data for weather prediction. You might have smart 91 00:07:55,959 --> 00:08:00,160 cities, you might have traffic sensors everywhere. And you know his point that, 92 00:08:00,480 --> 00:08:03,839 yes, there are sort of big computers, thirty two bit sixty four 93 00:08:03,879 --> 00:08:11,240 bit processors. In these systems of you know, thousands, tens of thousands, 94 00:08:11,319 --> 00:08:16,199 million, sometimes of devices, there's very few of those sort of big 95 00:08:16,319 --> 00:08:22,879 conventional computers analyzing the data. Most of what we're talking about here are very 96 00:08:22,879 --> 00:08:28,759 small. He's talked about eight bit and sixteen bit, talking about compute capability 97 00:08:28,360 --> 00:08:31,639 in a lot of these circumstances out in the middle of nowhere, there is 98 00:08:31,680 --> 00:08:35,159 no electric power. You're talking solar powered devices. You're talking not just limited 99 00:08:35,240 --> 00:08:41,440 compute power, but limited electrical power. This is sort of the big picture 100 00:08:41,480 --> 00:08:45,039 of the problem that Christopher's going to be talking about. And you know, 101 00:08:45,080 --> 00:08:48,720 when you put it like you just did, it occurs to me. You 102 00:08:48,759 --> 00:08:54,960 know, on this podcast we're usually not exclusively, but usually talking about you 103 00:08:56,000 --> 00:09:01,639 know, manufacturing, plants, refineries, hospital These are sites where you sort 104 00:09:01,679 --> 00:09:05,240 of have a boundary around them, for better or worse, and most of 105 00:09:05,279 --> 00:09:09,240 it is indoors. Maybe not entirely. It's like a controlled space. But 106 00:09:09,720 --> 00:09:13,399 as you just mentioned, you know, when you're talking about smart meters or 107 00:09:13,440 --> 00:09:18,759 traffic sensors or what have you, these are really widespread outdoor, sort of 108 00:09:18,840 --> 00:09:24,559 everywhere situations. I would wonder how this affects the security problem. And I'm 109 00:09:24,559 --> 00:09:30,240 sure that you guys are going to get into that. The main problem I 110 00:09:30,279 --> 00:09:35,360 heard you describe here was key management. And you know I've heard encryption described 111 00:09:35,480 --> 00:09:41,559 as a tool that turns every problem into the world into a key management problem. 112 00:09:41,759 --> 00:09:46,080 What is the key management problem when you're talking about I don't know, 113 00:09:46,240 --> 00:09:50,200 traffic sensors or smart meters. What can you give me an example? What's 114 00:09:50,279 --> 00:09:54,360 the problem? Why do we need a solution at all? I said cryptographic 115 00:09:54,440 --> 00:09:58,000 key management. Some people might call it trust management. Root of trust management. 116 00:09:58,279 --> 00:10:03,639 It's the piece this is in the center of your every device. And 117 00:10:03,840 --> 00:10:07,840 if you have an energy meter, for example, and a company makes a 118 00:10:07,960 --> 00:10:13,720 model and then they mass produce a million of them, the software is exactly 119 00:10:13,759 --> 00:10:18,399 the same. The only thing different is the serial number in there that tells 120 00:10:18,399 --> 00:10:24,759 it which one am I. But now in the age of virtualization, you 121 00:10:24,799 --> 00:10:30,360 can have a virtual machine running a simulated software meter a real software meter, 122 00:10:30,559 --> 00:10:35,159 and if they're all on the internet, you can't tell the difference. So 123 00:10:35,159 --> 00:10:41,279 some people might say the root of trust is an identification problem, and each 124 00:10:41,399 --> 00:10:45,919 device being able to identify itself, but to also uniquely say that I can 125 00:10:46,279 --> 00:10:50,039 not only identify who I am, but I can prove that I made this 126 00:10:50,159 --> 00:10:56,120 operation, I produce this data, I communicated with this other device, and 127 00:10:56,159 --> 00:11:00,480 the two devices can prove each other and say I'm actually talking to the device 128 00:11:00,639 --> 00:11:05,360 I'm supposed to be talking to, and not a virtual device that's mimicked of 129 00:11:05,399 --> 00:11:11,279 it, or some other device or somebody is in between listening all those pieces 130 00:11:11,679 --> 00:11:16,039 in security that make each device. And I know this gets kind of in 131 00:11:16,080 --> 00:11:20,960 the conceptional level, but we move it down to an energy meter. Can 132 00:11:22,000 --> 00:11:26,759 all my energy meters on my grid verify that I'm talking to them from my 133 00:11:26,919 --> 00:11:31,559 reading station, verify that when they're talking to they're talking to each other when 134 00:11:31,559 --> 00:11:37,399 we're collecting the data, and that the data is accurate and naturally came from 135 00:11:37,480 --> 00:11:43,960 them, so that we can then have valid information for billing, or valid 136 00:11:45,000 --> 00:11:50,399 information for our control structure or for our demand response. But all of the 137 00:11:50,240 --> 00:11:56,440 information that comes to and from any embedded system that people might not touch for 138 00:11:56,559 --> 00:12:01,639 ten years becomes questionable, whether it's authentic, or whether it's valid, or 139 00:12:01,679 --> 00:12:07,360 whether we're even talking to the advice we are with the interconnected age? What 140 00:12:07,480 --> 00:12:11,600 is the risk here? Let's say that I have you know, found on 141 00:12:11,639 --> 00:12:15,799 the internet, and someone has stolen a copy of the firmware for the smart 142 00:12:15,840 --> 00:12:18,440 meter that you know is in my home. I put it up in a 143 00:12:18,519 --> 00:12:22,440 virtual machine and now I'm impersonating a smart meter. I you know, in 144 00:12:22,639 --> 00:12:28,799 principle, can find the serial number on my device because it's they're attached to 145 00:12:28,840 --> 00:12:33,399 my home, and I, you know, can can you know embed it 146 00:12:33,440 --> 00:12:37,240 in the virtual machine? And now I can you know, disable somehow put 147 00:12:37,240 --> 00:12:41,200 you I don't know, tinfoil around the meter. U disabled communications between the 148 00:12:41,200 --> 00:12:43,679 meter and the grid, and now I can impersonate the meter and say, 149 00:12:43,720 --> 00:12:48,720 Andrew's not using any power? Is that what we're talking about? Is that 150 00:12:48,519 --> 00:12:52,080 the attack scenario or you know, is there something else you're thinking about? 151 00:12:52,759 --> 00:12:56,559 So when you're looking at a meter, the attack scenario is the company protecting 152 00:12:56,600 --> 00:13:00,639 itself from the customer. And I think that's kind of what you were alluding 153 00:13:00,639 --> 00:13:05,360 to. When we get into the day and age that everybody's a producer and 154 00:13:05,360 --> 00:13:11,039 a consumer and micro grids and powers coming onto and off the grid by different 155 00:13:11,039 --> 00:13:15,600 people at different times of the day, different different providers, the meter becomes 156 00:13:16,000 --> 00:13:20,639 your proof of whether you took power or provided power. To the grid. 157 00:13:20,360 --> 00:13:24,679 And if you, as a utility company, have a whole bunch of customers 158 00:13:24,679 --> 00:13:28,879 out there that are telling you they're providing power to your grid, and you 159 00:13:30,080 --> 00:13:35,519 now are expected to pay them, how do you prove that they're accurate? 160 00:13:35,679 --> 00:13:41,519 Is the question we're answering here In that case. In the it's not as 161 00:13:41,559 --> 00:13:46,000 important if all the meters are owned by the company other than to get an 162 00:13:46,039 --> 00:13:50,919 overall we're not getting falsified. But when we get into that now renewable peer 163 00:13:50,919 --> 00:13:58,120 to peer energy ecosystem of micro grids, we can't trust the people. If 164 00:13:58,159 --> 00:14:05,679 we can't trust the peaceeople providing the information from a microgrid to make monetary transactions 165 00:14:05,679 --> 00:14:11,720 based off of we end up with anybody being able to scam the system. 166 00:14:11,159 --> 00:14:16,120 And definitely because you can't prove that they put power on other than their communication, 167 00:14:16,799 --> 00:14:22,879 which of their communication is not provable mathematically, cryptographically. And that's what 168 00:14:22,320 --> 00:14:26,799 this root of trust in the distributed devices gives you. You know, to 169 00:14:26,840 --> 00:14:31,720 me, clearly, there's there's a need. You know, you've convinced me 170 00:14:31,759 --> 00:14:35,039 there's a need to identify these meters, especially if they're putting power onto the 171 00:14:35,080 --> 00:14:39,200 grid. I mean partly billing, Yes, but partly, if you've got 172 00:14:39,320 --> 00:14:45,159 enough of these things, there's grid stability that might be at risk. But 173 00:14:45,559 --> 00:14:50,480 is the solution not that when you know, the technician shows up at my 174 00:14:50,600 --> 00:14:54,720 home to attach a meter to my house. Is the solution not that the 175 00:14:54,799 --> 00:15:00,320 technician writes down the serial number and you know, presumably has a there's a 176 00:15:00,360 --> 00:15:05,279 database somewhere associating serial numbers with private keys that are built into the into the 177 00:15:05,600 --> 00:15:09,159 device. Is that not the solution? Why do we need something more complex? 178 00:15:11,559 --> 00:15:13,279 So the solution, you know, if you think about what you just 179 00:15:13,320 --> 00:15:20,679 said, where the private key of the device, right, And you're the 180 00:15:20,360 --> 00:15:26,039 solution set isn't the private key in the device. It's getting the private key 181 00:15:26,080 --> 00:15:33,600 to the device. And if that's somebody doing it every time a vendor puts 182 00:15:33,600 --> 00:15:37,000 one on the grid, there's fifty different vendors that makes products for it. 183 00:15:37,720 --> 00:15:43,919 Everybody does it differently, and everybody has a different certificate authority if they're using 184 00:15:43,919 --> 00:15:48,519 certificates, and the certificates can be changed, they can be stolen, they 185 00:15:48,519 --> 00:15:54,360 can be spoofed. In a virtual machine, you can imitate another machine entirely. 186 00:15:56,080 --> 00:15:58,559 So if we scale back and step back and say, yeah, the 187 00:15:58,679 --> 00:16:03,440 meter set is one problem in the consumer base, and we're moving towards that 188 00:16:03,559 --> 00:16:11,399 distributed consumer producer environment. The bigger problem, at least the now problem, 189 00:16:11,519 --> 00:16:15,799 is in the control structure, the demand response and the infrastructure of control, 190 00:16:15,519 --> 00:16:22,159 and that they have equipment made by many different vendors that go into this and 191 00:16:22,200 --> 00:16:27,679 they have to have microsecond responses from trusted equipment, and the only way to 192 00:16:27,679 --> 00:16:33,000 do that is to keep it on a private grid, private network. And 193 00:16:33,080 --> 00:16:38,120 each utility company now wants to start interconnecting and working across utility companies. They 194 00:16:38,120 --> 00:16:42,559 have hundreds of thousands of devices that a human might not touch for five or 195 00:16:42,559 --> 00:16:52,120 ten years and keep keeping and maintaining these those keys that you just talked about, 196 00:16:52,159 --> 00:16:56,120 the private keys in every device is a human centric problem, and that's 197 00:16:56,159 --> 00:17:03,759 where we're falling short. What we're talking talking about here is modularizing and making 198 00:17:03,840 --> 00:17:10,960 that human solution done uniformly to have their cryptographic keys managed so you can see 199 00:17:10,960 --> 00:17:15,599 them not just for a grid, not just for a utility company, but 200 00:17:15,680 --> 00:17:19,359 for the industry many different companies, many different vendors that all come in and 201 00:17:19,359 --> 00:17:25,559 put products in this kind of Frankenstein mesh, and all of them are different, 202 00:17:25,640 --> 00:17:27,680 and all of them are done by different people who we might not know, 203 00:17:27,799 --> 00:17:34,400 might not trust, different levels of software, different levels of responsible people. 204 00:17:34,559 --> 00:17:40,119 Even maybe something made in a supply chain where it was in another country 205 00:17:40,160 --> 00:17:45,839 and there's purposeful, harmful information or malware injected into it. So that mish 206 00:17:47,839 --> 00:17:55,039 mess of unpredictableness is the problem that really is plaguing that whole industry. We 207 00:17:55,759 --> 00:18:03,279 are all working in our own space without the to know that who we're talking 208 00:18:03,319 --> 00:18:08,319 to and what we're putting on our networks and into our systems is actually authentic. 209 00:18:10,400 --> 00:18:11,799 Okay, So you know that kind of frames the problem for me. 210 00:18:11,839 --> 00:18:17,359 We're talking millions of devices on the grid. Some of them are measuring power 211 00:18:17,400 --> 00:18:21,000 consumption, some of them are measuring power productions, some of them are measuring 212 00:18:21,000 --> 00:18:23,920 other things. There's a lot to measure when we're when we're talking about that 213 00:18:25,200 --> 00:18:30,319 level of granularity. And yet you know, while I take your point that 214 00:18:30,920 --> 00:18:34,279 you know there can be manual intervention in the course of deploying the equipment. 215 00:18:34,920 --> 00:18:40,039 It could be from lots of different vendors, and if it's going to sit 216 00:18:40,079 --> 00:18:42,160 there for five or ten or fifteen years, the way this stuff often does, 217 00:18:42,200 --> 00:18:45,440 I mean, you know, when was the last time you updated the 218 00:18:45,440 --> 00:18:48,680 firmware, and you're a refrigerator if it's going to sit there for a long 219 00:18:48,720 --> 00:18:53,480 time. These keys age best practices that, especially on the Internet, you 220 00:18:53,519 --> 00:18:56,920 don't leave the key in there forever. There's got to be a way to 221 00:18:56,039 --> 00:19:02,640 update these and you know, coordinate the updates across producers and consumers of the 222 00:19:02,640 --> 00:19:07,799 information. It all makes sense that this is a problem. You've got a 223 00:19:07,839 --> 00:19:11,400 technology. Can you talk about the technology? What kind of technology are you 224 00:19:11,440 --> 00:19:18,160 proposing as the solution to this problem. Yeah, So from an industry level, 225 00:19:18,559 --> 00:19:22,720 working with many different vendors and literally fifteen years, twenty years of my 226 00:19:22,839 --> 00:19:27,440 career has been working to solve this problem is identifying what are all the things 227 00:19:27,480 --> 00:19:32,200 that they have been doing. And the first part was to create a modular 228 00:19:33,079 --> 00:19:37,079 piece that could go into any device that would handle all the things that everybody 229 00:19:37,160 --> 00:19:44,240 is using for security and simplifying it. The interesting thing is we overcomplicate what 230 00:19:45,640 --> 00:19:49,759 the is being done and we make a thousand different applications. But security is 231 00:19:49,839 --> 00:19:52,960 kind of like an art form as much of it as it is a technology. 232 00:19:53,279 --> 00:19:59,079 Where in artwork we only have nine components that make up paintings and pictures 233 00:19:59,079 --> 00:20:03,759 and sculptures and artists learn these in college and once you can master each one 234 00:20:03,759 --> 00:20:10,039 of the nine components like line, form, volume, shape, parallax. 235 00:20:10,519 --> 00:20:15,440 Uh, there's there's nine of them. Cybersecurity similarly, there's only seven things 236 00:20:15,440 --> 00:20:22,640 we do, and people might challenge this, but we identify data, we 237 00:20:22,680 --> 00:20:27,400 authenticate users and systems, we establish connections, we hide data or encrypt it, 238 00:20:29,440 --> 00:20:36,240 We blog and operations and verify that the the data and the operations, 239 00:20:36,480 --> 00:20:41,000 and we distribute trust. I think that was seven. So if we make 240 00:20:41,039 --> 00:20:45,359 a way to unify the ability to do all those, we can put that 241 00:20:45,519 --> 00:20:51,960 modular approach into every system, every vendor, every to every product, and 242 00:20:52,000 --> 00:20:56,880 then we the second thing is to make a way to change that out because 243 00:20:56,880 --> 00:21:00,279 we know that we can't keep the same security forever. So the second part 244 00:21:00,319 --> 00:21:06,079 of what we do is a method using distributed ledger known as blockchain, to 245 00:21:06,119 --> 00:21:11,039 be able to change out those credentials, the human piece that we usually do 246 00:21:11,160 --> 00:21:14,240 with a guy running in the truck to it, and be able to do 247 00:21:14,279 --> 00:21:18,119 that and prove that it was done globally so everybody sees it, so we 248 00:21:18,160 --> 00:21:26,079 can prove who it basically is responsible for that provisioning. It's called of those 249 00:21:26,119 --> 00:21:30,759 cryptographic components, and it doesn't have to be somebody that you don't trust. 250 00:21:30,839 --> 00:21:34,440 It's open. And this is where we kind of borrowed that distribution of trust 251 00:21:34,880 --> 00:21:40,759 using blockchain what we know from like cryptocurrency, but using it in a totally 252 00:21:40,799 --> 00:21:47,759 different way to actually distribute the root of trust the cryptographic keys, which are 253 00:21:48,039 --> 00:21:53,079 all standard cryptography. Just doing the human aspect, but proving that, improving 254 00:21:53,559 --> 00:21:59,799 the providence of it, the attestation of how that happened over your whole grid, 255 00:21:59,799 --> 00:22:03,519 over every product, or over your supply chain throughout the life cycle of 256 00:22:03,519 --> 00:22:10,480 the product, and maintaining that over time an over distributed area and over geographic 257 00:22:10,640 --> 00:22:15,440 and over logistics of networks, and even the people aspect of it, connecting 258 00:22:15,559 --> 00:22:22,519 your workforce and maintenance. Andrew, we've done somewhere north of one hundred episodes 259 00:22:22,519 --> 00:22:26,960 of this podcast at this point. Are there really only seven steps to industrial 260 00:22:27,000 --> 00:22:30,359 security? Here? I should have asked, I was. I was kind 261 00:22:30,400 --> 00:22:34,279 of wondering that myself. I did it quild Google afterwards. I haven't seen 262 00:22:34,319 --> 00:22:38,759 where the seven steps come from. But you know, if you ask me, 263 00:22:38,880 --> 00:22:45,160 those are seven steps that are integral to communications security, and you know, 264 00:22:45,319 --> 00:22:48,400 yeah, we've been doing one hundred episodes. We're talking about the big 265 00:22:48,440 --> 00:22:52,119 picture is more than communications. It has to do with physical security, It 266 00:22:52,119 --> 00:22:56,359 has to do with you know, people, process and technology, has to 267 00:22:56,400 --> 00:23:00,599 do with with you know, host based a lot host based stuff, you 268 00:23:00,640 --> 00:23:07,559 know as well, you know, concrete example, if the you know, 269 00:23:07,599 --> 00:23:11,000 I've got a host, I've got you know, it's a it's a I 270 00:23:11,039 --> 00:23:15,079 don't know, a Windows box, a Linux box, a server sixty four 271 00:23:15,119 --> 00:23:18,839 bits, big, big operating system. And it turns out that my crypto 272 00:23:18,920 --> 00:23:25,240 library has a vulnerability in it, and that vulnerability can be exploited simply by 273 00:23:25,279 --> 00:23:30,680 sending a message across the Internet into the machine, into the host, into 274 00:23:30,680 --> 00:23:34,480 the server, and that compromises the library. It makes it you know, 275 00:23:34,559 --> 00:23:38,119 I don't know, buffer overflow, it makes it do bad things. You 276 00:23:38,160 --> 00:23:42,960 know. No amount of key management is going to solve that vulnerability problem. 277 00:23:44,000 --> 00:23:48,599 That's a patch problem. So if we're talking about pushing data across the wire, 278 00:23:48,920 --> 00:23:52,720 if the receiver the server is asking the question can I trust the data? 279 00:23:52,799 --> 00:23:56,279 Well, then the seven things that Christopher's talking about here, these are 280 00:23:56,319 --> 00:24:00,200 all steps that we do have to have in place. But the bigger picture 281 00:24:00,279 --> 00:24:06,599 is there's a little bit more to it. So that makes sense. I 282 00:24:06,640 --> 00:24:07,920 mean, we need to change these keys from time to time. We've got 283 00:24:07,960 --> 00:24:15,680 lots of different vendors and you know, power utilities and others involved. When 284 00:24:15,759 --> 00:24:19,119 we make a change to the key somewhere, we have to publish that. 285 00:24:19,119 --> 00:24:25,640 That publication mechanism has to be standards so that everybody can consume the knowledge that 286 00:24:25,680 --> 00:24:27,440 we've just changed the key. Everyone who has to communicate with this device. 287 00:24:29,200 --> 00:24:32,039 We have to make sure that that process is authentic, that it can't be 288 00:24:32,079 --> 00:24:34,920 spoofed by someone trying to you know, steal power or you know, do 289 00:24:36,000 --> 00:24:40,960 other malicious things. And the solution that you're proposing is blockchain. So can 290 00:24:41,000 --> 00:24:45,720 we talk about the solution in a bit more detail. I mean bitcoin is 291 00:24:45,880 --> 00:24:52,519 power hungry. Bitcoin, you know, there's farms of servers involved. It's 292 00:24:52,599 --> 00:24:56,720 not going to work with an eight bit CPU. What what does your solution 293 00:24:56,839 --> 00:25:02,680 really look like? There's a lot behind that simple question, and it comes 294 00:25:02,799 --> 00:25:07,359 you know, in two parts. One, the part that we're provisioning is 295 00:25:07,480 --> 00:25:12,279 very tiny and literally the cryptographic keys and the continued operation only takes up sixty 296 00:25:12,359 --> 00:25:17,960 four K of memory in a small sensor device. So now we look at 297 00:25:18,079 --> 00:25:21,279 that's what goes out, and that's the key distribution. That's what goes on 298 00:25:21,319 --> 00:25:25,759 your device. It doesn't take much size. Now, the distribution of blockchain 299 00:25:25,839 --> 00:25:29,799 was a whole other problem set. This actually spent the last six years my 300 00:25:29,920 --> 00:25:36,559 PhD dissertation in solving the problem of scaling blockchain, and we actually throughout traditional 301 00:25:36,720 --> 00:25:40,559 blockchain. That's why I always try to say we have a distributed ledger. 302 00:25:40,880 --> 00:25:45,440 We now are going to the third generation. Of the first generation was cryptocurrency, 303 00:25:45,599 --> 00:25:51,119 the second generation was smart contracts and your ethereum, your hyperledger fabric. 304 00:25:51,400 --> 00:25:56,680 And now we have a third generation that doesn't use the mining, a different 305 00:25:56,839 --> 00:26:03,839 algorithm that turns around the word to be your time spent participating storing data over 306 00:26:03,880 --> 00:26:08,680 time is how you gain your incentive versus a wasted energy up front. And 307 00:26:08,759 --> 00:26:15,200 we're actually running peer ledger distributed ledger nodes on a Raspberry Pie that's how small 308 00:26:15,359 --> 00:26:19,039 they can run on. And then there's a it's designed modularly so it can 309 00:26:19,079 --> 00:26:25,200 expand indefinitely, where we have a consortium that's similar to your dns on the 310 00:26:25,240 --> 00:26:29,519 Internet, that manages who's out there and the governance of it. And then 311 00:26:29,599 --> 00:26:34,160 each individual peer that can scale horizontally like your routers on the Internet, and 312 00:26:34,200 --> 00:26:41,279 we're turning into this new animal of no mining and indefinite scalability. I think 313 00:26:41,319 --> 00:26:44,319 it's you know, I'm excited about it, but it's been my six seven 314 00:26:44,400 --> 00:26:48,759 years of work, and you know, I love to help anybody understand more 315 00:26:48,799 --> 00:26:52,279 about it. Bring in ask questions as much question as if you want. 316 00:26:52,759 --> 00:26:59,160 My dissertation is out publication. You can find it under a sustainable framework for 317 00:26:59,200 --> 00:27:02,799 a distributed ledger. There is the title of it on ProQuest and it's open 318 00:27:02,839 --> 00:27:07,240 for anybody to view. Let me dig just little detail deeper. You know, 319 00:27:07,279 --> 00:27:11,599 you talked about the memory footprint in the device. You talked about the 320 00:27:12,680 --> 00:27:18,119 you know, the Raspberry pie is managing the ledger. Something you didn't mention 321 00:27:18,359 --> 00:27:21,440 is the size of the ledger. I mean, in my understanding, the 322 00:27:21,440 --> 00:27:23,680 Bitcoin ledger is now I don't know, I don't know, like a dozen 323 00:27:23,759 --> 00:27:30,480 terabytes or something like that. It's and and that's you know, that's money 324 00:27:30,519 --> 00:27:34,680 changing hands. If we have you know, millions of devices coming out of 325 00:27:34,680 --> 00:27:40,960 each of dozens of vendors going into you know, hundreds of power utilities all 326 00:27:40,960 --> 00:27:44,400 over the world, most of which are connected to at least one other power 327 00:27:44,519 --> 00:27:48,680 utility. There's there's very few power utilities that operate in complete isolation. It 328 00:27:49,039 --> 00:27:53,039 seems to me that, you know, is it the case Let me ask 329 00:27:53,119 --> 00:27:57,640 you, is it the case that you have to be able to share all 330 00:27:57,720 --> 00:28:03,160 of this key information for every device on the planet with every possible consumer on 331 00:28:03,200 --> 00:28:07,920 the planet. And if so, is that going to scale sort of storage 332 00:28:07,960 --> 00:28:14,640 wise? And that was actually the problem we set out to solve because whenever 333 00:28:15,079 --> 00:28:19,720 we came up with the modular approach to distributing cryptographic keys, we actually tried 334 00:28:19,759 --> 00:28:26,319 it on some second generation blockchains and used hyper ledger, fabric, sawtooth Lake 335 00:28:26,759 --> 00:28:30,519 and ethereum, and it brought it to its knees. There's no way either 336 00:28:30,519 --> 00:28:34,240 any of them could scale. So the whole design of a scalable blockchain was 337 00:28:34,279 --> 00:28:40,440 based on being able to hit the mark of the needs for cryptographic key management 338 00:28:40,559 --> 00:28:45,720 and how big that would get. And basically it's we came up with a 339 00:28:45,839 --> 00:28:52,279 loosely coupled chain of chains, so side chains that operate independently. And that's 340 00:28:52,279 --> 00:28:55,119 what I said, like a router, Like a set of routers you plug 341 00:28:55,319 --> 00:28:59,960 in and they all work together, but not everybody has to have everybody's day 342 00:29:00,039 --> 00:29:03,759 data. There's sub segments and it makes it scalable, but yet it's the 343 00:29:03,799 --> 00:29:11,960 same forensic mapping to prove everybody's transactions and timestamp them globally. And it's a 344 00:29:11,119 --> 00:29:18,000 unique consensus approach called a synchronous trust consensus model, where it adds a couple 345 00:29:18,039 --> 00:29:22,759 of major things. One that we no longer have to keep data all the 346 00:29:22,759 --> 00:29:27,599 way back to the genesis block. The data is kept as long as it 347 00:29:27,720 --> 00:29:33,400 has value to the operator and until you get your incentive, which is set 348 00:29:33,440 --> 00:29:37,839 for the longest time in about fourteen years, but only in subsets of data 349 00:29:37,880 --> 00:29:41,319 as well, so not everybody has everybody's data, and the consortium service keep 350 00:29:41,359 --> 00:29:48,119 the validation data of all the individual groups of ledgers that can now scale indefinitely 351 00:29:48,519 --> 00:29:53,400 across the globe. My testing and you can read it in my dissertation where 352 00:29:53,759 --> 00:30:00,720 we scaled this up to fifty two million ledgers, which is currently the size 353 00:30:00,720 --> 00:30:03,640 of the number of routers on the Internet, to see if it would if 354 00:30:03,640 --> 00:30:08,920 it would still be feasible and operational, and the scalability models predicted that we 355 00:30:08,960 --> 00:30:15,359 could do this and that the growth in size and even the data we're going 356 00:30:15,400 --> 00:30:19,920 to persist over time is manageable because we can phase it out over time and 357 00:30:19,960 --> 00:30:26,839 basically have a window moving forward. And that management of data and the ability 358 00:30:26,880 --> 00:30:33,200 to meet this mark is and has been that primary focus. So kind of 359 00:30:33,240 --> 00:30:36,720 excited, like once again, open and talk more about this, and you'll 360 00:30:36,759 --> 00:30:41,440 hear me speaking all over the place on it. But it is really a 361 00:30:41,559 --> 00:30:45,920 very involved topic. We've had over four hundred people in the state of Colorado 362 00:30:47,000 --> 00:30:52,960 involved, over eighty five people turn code on community source projects to and this 363 00:30:52,079 --> 00:30:56,720 has been funded by the State of Colorado under legislation we wrote in twenty eighteen 364 00:30:57,279 --> 00:31:00,759 through the University of Colorado, colors University, a couple others. And that's 365 00:31:00,759 --> 00:31:04,079 why it's so big and exciting, because we've had so many people work on 366 00:31:04,160 --> 00:31:08,160 it and getting excited about it that this is actually something that we can see 367 00:31:08,400 --> 00:31:14,440 scaling to that next generation for solving those problems you brought up in your question, 368 00:31:15,480 --> 00:31:19,319 so real quick, just a clarifying question. You know, you suggested 369 00:31:19,359 --> 00:31:23,359 fifty two million routers, you said fifty two million ledgers. Does that mean 370 00:31:23,440 --> 00:31:29,240 that each ledger could in turn manage thousands or I don't know, a million 371 00:31:29,279 --> 00:31:33,279 devices, And so that's the you know, the fifty two million times one 372 00:31:33,279 --> 00:31:37,039 thousand or a million that we're talking about, or is each leg does is 373 00:31:37,079 --> 00:31:41,119 there a ledger per device? When I'm talking fifty two million ledgers, I'm 374 00:31:41,160 --> 00:31:45,599 talking to the number of ledgers that can store an indefinite amount of blocks as 375 00:31:45,640 --> 00:31:52,519 many as they can handle on the blockchain, So the amount of data transactions 376 00:31:52,920 --> 00:32:00,240 is astronomically larger than that. Admittedly, Andrew, whenever this kind of subject 377 00:32:00,240 --> 00:32:05,400 comes up, I become a little bit suspicious. If I had a nickel 378 00:32:05,480 --> 00:32:08,599 for every time somebody sold me on a scalable blockchain, I probably wouldn't have 379 00:32:08,680 --> 00:32:13,960 to do podcasts anymore. So. The question that I suppose I would ask 380 00:32:14,000 --> 00:32:19,759 of doctor Gorrogh, which may well be answered in his dissertation or later in 381 00:32:19,799 --> 00:32:24,119 your interview here, is whether this kind of blockchain solution that he is describing, 382 00:32:25,240 --> 00:32:30,680 while scalable and fast and whatever usable as you need, offers the same 383 00:32:30,839 --> 00:32:36,799 kinds of security protections or doesn't compromise too much on them compared with the other 384 00:32:36,839 --> 00:32:39,960 blockchains that we're using as comparisons, you know, Bitcoin ethereum, the ones 385 00:32:40,000 --> 00:32:46,400 that are slow and unscalable for reasons that aren't trivial good question. I did 386 00:32:46,400 --> 00:32:52,400 not ask that question. I do know that, you know, one difference 387 00:32:52,480 --> 00:33:00,599 between the system that doctor gorrog is talking about and the traditional sort of bitcoin 388 00:33:00,680 --> 00:33:07,039 blockchain is power usage. I mean, the bitcoin blockchain already uses a measurable 389 00:33:07,039 --> 00:33:09,960 fraction of all of the world's power, and we're talking about you know, 390 00:33:10,039 --> 00:33:15,200 a tiny, tiny fraction of all the world's computers involved in that blockchain, 391 00:33:15,599 --> 00:33:21,160 whereas you know, here we're talking about eight and sixteen bit devices, millions 392 00:33:21,200 --> 00:33:28,400 of them in every city involved in this ledger. The second sort of thing 393 00:33:28,480 --> 00:33:30,359 is I don't even know if this is the right question to be asking. 394 00:33:30,799 --> 00:33:36,160 You know, the bitcoin blockchain models the movement of money, whereas here this 395 00:33:36,240 --> 00:33:39,960 blockchain models trusted models. You know, the degree to which we can trust 396 00:33:40,160 --> 00:33:45,000 different devices within an organization, across organizations. I don't even know if these 397 00:33:45,039 --> 00:33:50,519 are the same questions. I mean, when we're talking about trusting things, 398 00:33:50,640 --> 00:33:52,279 the thing that springs to mind, the system that springs to mind is active 399 00:33:52,279 --> 00:33:59,319 directory. It's the classic system that's used for managing users, not even devices. 400 00:33:59,359 --> 00:34:01,880 Actric directory doesn't manage devices, to my knowledge, to manages users, 401 00:34:02,359 --> 00:34:07,760 and we're talking about, you know, a system that could in principle scale 402 00:34:07,800 --> 00:34:15,360 to all of the devices on the planet being interconnected to some degree, and 403 00:34:15,840 --> 00:34:20,760 I don't know how big active direct real scale, but I'll be deeply surprised 404 00:34:20,800 --> 00:34:23,440 if it scales to all the users on the planet, much less the one 405 00:34:23,480 --> 00:34:30,159 hundred devices per user that we're expecting to see deployed in the next century. 406 00:34:30,639 --> 00:34:34,960 So firnt answer is, I don't know. The longer answer is, I 407 00:34:35,000 --> 00:34:38,239 think it has to do with power usage, with scalability to you know, 408 00:34:38,320 --> 00:34:46,000 sort of ridiculous scales. Is my understanding. You folks that you've personally been 409 00:34:46,039 --> 00:34:50,440 involved in this for a long time. Block Frame as a company has been 410 00:34:50,480 --> 00:34:53,880 involved in it for a long time. Can you talk about block Frame a 411 00:34:53,880 --> 00:34:57,840 little bit more? What have you got in this space? You know, 412 00:34:57,880 --> 00:35:00,400 if people want if a vendor called up and said, hey, you know, 413 00:35:00,480 --> 00:35:05,239 I want to do this, you've convinced me. Do you have technology? 414 00:35:05,239 --> 00:35:08,840 What are you offering in the space? Yeah, so we write block 415 00:35:08,880 --> 00:35:14,920 frame has been around for seven years. The technology from my dissertation is wrapped 416 00:35:14,920 --> 00:35:19,840 into Blockframe. We have literally patented the consensus model where we'll be the only 417 00:35:19,880 --> 00:35:25,840 ones globally able to implement this as well as opened up a community project where 418 00:35:27,000 --> 00:35:30,800 a lot of this work was done for the State of Colorado, funded under 419 00:35:30,880 --> 00:35:36,400 many different sources, and a big piece of it is open source interfaces that 420 00:35:36,480 --> 00:35:40,480 anybody can develop on top of to make your community your project, for your 421 00:35:40,519 --> 00:35:46,679 application, write your own smart contract, do your own embedded application that implements 422 00:35:46,719 --> 00:35:53,360 on top of the distributed ledgers for the blockchain side, or we also Blockframe 423 00:35:53,519 --> 00:36:00,559 offers the modular ability and whether a hardware insert a physical device that goes on 424 00:36:00,599 --> 00:36:07,760 your your IoT device that holds your cryptographic keys, or a modular We even 425 00:36:07,760 --> 00:36:10,280 have a software approach. The hardware is always going to be more secure that 426 00:36:10,360 --> 00:36:15,280 we can implement. Work with vendors right now to implement this and get into 427 00:36:15,320 --> 00:36:20,039 your design for your next release, or work on getting it into patches that 428 00:36:20,079 --> 00:36:23,400 could go out in legacy systems, so we can. We're offering for one 429 00:36:23,599 --> 00:36:29,960 cost through the release or your product to get you on board, gets you 430 00:36:30,280 --> 00:36:35,079 all your your tokenization on the blockchain, all your embedded designs, set up 431 00:36:35,480 --> 00:36:39,000 support with making getting your product to markets, and we offer that all to 432 00:36:39,199 --> 00:36:44,679 any vendor who wants to be an early adopter on this. If you're looking 433 00:36:44,760 --> 00:36:47,960 at hey, I want to be on the next generation of security that's not 434 00:36:49,079 --> 00:36:53,679 only modular, but peer to peer based. Now every device can go unique 435 00:36:53,800 --> 00:36:59,559 one to each other, which opens up that thing that we've been looking for 436 00:36:59,599 --> 00:37:04,320 and why people got into blockchain so heavily, because it opens up that peer 437 00:37:04,320 --> 00:37:07,719 to peer, it takes out the middleman. It makes it so now every 438 00:37:07,880 --> 00:37:14,280 one of our devices are individualized and can have a verification of each other. 439 00:37:14,719 --> 00:37:19,679 Zero trust between any two devices because they can identify each other on the blockchain 440 00:37:20,079 --> 00:37:25,559 before they communicate or commit to any operations and then do trusted proof of origins 441 00:37:25,599 --> 00:37:30,840 signed data between the two of them. So this the the level of security 442 00:37:30,920 --> 00:37:36,559 is something we haven't seen in our day and age where we're looking at you 443 00:37:36,559 --> 00:37:40,440 know, to date we've collectivized data, all data is held by a company 444 00:37:40,480 --> 00:37:45,320 an organization. We're looking at this migration to end, node to the edge, 445 00:37:45,360 --> 00:37:50,159 to peer to peer, and it's just a new paradigm and how we 446 00:37:50,199 --> 00:37:53,559 will look at security in the future. One other thing that struck me in 447 00:37:53,840 --> 00:37:59,239 your description of the solution. You know, there's technology involved, there's you 448 00:37:59,239 --> 00:38:04,800 know, sort of a community caations infrastructure, the distributed ledger involved. You 449 00:38:04,880 --> 00:38:13,000 mentioned that there's patents involved. How how interoperable is this If you know, 450 00:38:13,519 --> 00:38:16,280 a group of vendors on the other side of the planet wanted to do this 451 00:38:16,400 --> 00:38:21,239 stuff, I don't know on their own, could they is you know, 452 00:38:22,239 --> 00:38:24,480 is there or you know, is there a sort of a standard of standards 453 00:38:24,519 --> 00:38:30,920 where these many different kinds of ledgers can interoperate. How I guess how universal? 454 00:38:31,159 --> 00:38:36,000 You know? How how university available is the data here? Yeah? 455 00:38:36,039 --> 00:38:39,280 And so we actually, like I said, we designed this for standardization, 456 00:38:39,400 --> 00:38:50,119 for modern modularization. So every single cryptographic root of trust in all the devices 457 00:38:50,280 --> 00:38:54,519 is a modular block Now what is in those is entirely different for each one 458 00:38:54,960 --> 00:39:00,599 and your vendor, your product owner, for all their decides on certain pieces 459 00:39:00,599 --> 00:39:05,440 of it. The industry utility registrar, which is a kind of like a 460 00:39:05,559 --> 00:39:13,360 DNS server that programs that provisions a set of products out there, is able 461 00:39:13,440 --> 00:39:20,519 to then uniquely identify each one. And we are licensing those registrars for different 462 00:39:20,559 --> 00:39:23,440 companies to use, for different industries to use. So not only is the 463 00:39:23,559 --> 00:39:30,760 individual pieces of it modular, the blockchain for the wayback machine, you can 464 00:39:30,039 --> 00:39:36,320 revert back to a earlier stage that you did trust the device if it becomes 465 00:39:36,400 --> 00:39:40,079 compromised and reprovisioned from that. But now we have the ability for multiple different 466 00:39:40,480 --> 00:39:46,239 people, multiple different product owners, to take charge of their segment of the 467 00:39:46,320 --> 00:39:54,840 market. And we have a we actually did a decentralized autonomous organization that owns 468 00:39:54,880 --> 00:40:00,119 the rights to run and manage the public distributed Ledger, and that's the State 469 00:40:00,159 --> 00:40:04,559 of Colorado, through the University of Colorado, is a part owner in that 470 00:40:05,000 --> 00:40:09,440 because many people provided that effort. So the whole thing is designed around these 471 00:40:09,559 --> 00:40:15,760 modular pieces that are developed for anybody to build on top. And then even 472 00:40:15,880 --> 00:40:21,199 the smart contracts then become ownership of the person that wrote them, and they 473 00:40:21,199 --> 00:40:25,320 can sublet or sell their smart contract for other people to use for whatever price 474 00:40:25,360 --> 00:40:30,800 they want. And you have all these modular constructs within the overall system of 475 00:40:30,880 --> 00:40:37,119 systems that makes everybody in the marketplace be able to run their own business and 476 00:40:37,719 --> 00:40:40,800 be individual. The control structures that we put together is just to make sure 477 00:40:40,800 --> 00:40:47,559 that we don't have competing standardization pieces, and that's why we decided to go 478 00:40:47,599 --> 00:40:51,920 with the patent route, so we didn't have somebody competing against releasing the same 479 00:40:51,920 --> 00:40:55,199 thing on the other side of the world. That we can then say from 480 00:40:55,239 --> 00:41:00,639 the first couple phases, we get it very standard before we start that kind 481 00:41:00,639 --> 00:41:07,639 of push and release structure. All right, Andrew, we are really getting 482 00:41:07,800 --> 00:41:12,639 in the weeds of blockchain stuff. And I'm only faintly remembering that we are 483 00:41:12,639 --> 00:41:16,320 talking about smart meters here. Can you help me with the connection? Yeah, 484 00:41:16,360 --> 00:41:21,480 so sure, you know, it's not it's more than smart meters. 485 00:41:21,519 --> 00:41:29,800 It's devices all over the place that need to talk to sort of control systems, 486 00:41:29,800 --> 00:41:32,840 skata systems, sort of you know, high end analytical systems all over 487 00:41:32,840 --> 00:41:37,559 the place. The example in the in the power grid is yes, smart 488 00:41:37,719 --> 00:41:42,159 smart meters is one piece of it. It's information really, but we're talking 489 00:41:42,159 --> 00:41:47,880 about about gathering information from many, many devices, and we're talking about sharing 490 00:41:49,159 --> 00:41:53,280 information and sometimes even sharing devices between organizations. So it's not just the meters. 491 00:41:53,840 --> 00:41:59,320 It's also the devices that are connected to you know, millions of rooftop 492 00:41:59,440 --> 00:42:04,280 solar that are sometimes producing power and sending it into the grid, and you 493 00:42:04,320 --> 00:42:08,280 know, the household is sometimes consuming power from the grid, and you know 494 00:42:08,400 --> 00:42:14,239 these organs who wants to talk to these devices Whilst yeah, the grid wants 495 00:42:14,280 --> 00:42:17,760 to talk to the devices the local utility, but sometimes there's other aggregators like, 496 00:42:17,880 --> 00:42:22,039 you know, I think Google has a system now where they can talk 497 00:42:22,079 --> 00:42:27,440 to your rooftop solar and aggregate your rooftop solar so that they can you know, 498 00:42:27,559 --> 00:42:30,440 interact sort of more aggressively, more personally, you know, with human 499 00:42:30,480 --> 00:42:37,920 oversight into power pricing to maximize the price that householders get for the power they 500 00:42:37,960 --> 00:42:43,920 give back into the grid to to and of course Google takes a cut of 501 00:42:44,519 --> 00:42:47,280 all that money. But you know, there we've got an example of a 502 00:42:47,320 --> 00:42:52,159 couple of organizations talking to the same rooftop solder. You've got synchrophasers all over 503 00:42:52,159 --> 00:42:55,320 the grid which are talking about, you know, measuring the in a sense, 504 00:42:55,400 --> 00:43:00,559 the health of the grid. It's you know, I won't go into 505 00:43:00,559 --> 00:43:04,039 phase measurement, but it's it's a technical thing that's done, and these synchrophaser 506 00:43:04,119 --> 00:43:08,480 measurements make sense to share across the many utilities that are cooperating in the grid. 507 00:43:08,880 --> 00:43:15,360 There's load shedding devices that have the ability to shed load that need you 508 00:43:15,400 --> 00:43:21,800 know, instruction from outfits like Google that are maximizing what you get paid for 509 00:43:21,840 --> 00:43:24,280 shedding the load that need to be you know, connected and report to the 510 00:43:24,320 --> 00:43:30,800 local utility and possibly other utilities. You know, you've got high voltage charging 511 00:43:30,840 --> 00:43:34,119 stations coming online everywhere that are pulling a lot of power from the grid that 512 00:43:34,239 --> 00:43:37,519 need to interact with the local utility and possibly bigger utilities. You know. 513 00:43:37,639 --> 00:43:43,519 In traffic, you've got different cities that are you know, coming up against 514 00:43:43,559 --> 00:43:45,880 each other, different jurisdictions. They might not want to know what's coming their 515 00:43:45,920 --> 00:43:50,679 way the simplest way and doing that might just be you know, connect to 516 00:43:50,760 --> 00:43:54,440 some of the other cities traffic sensors. But now you've got multiple organizations, 517 00:43:54,480 --> 00:44:00,719 multiple cities talking to the same sensor. You know, it's all about devices 518 00:44:00,800 --> 00:44:02,400 that need to talk to each other, that need to talk to a central 519 00:44:02,440 --> 00:44:09,360 analytical station, and where you've got multiple jurisdictions that need to trust, you 520 00:44:09,400 --> 00:44:15,239 know, that might need to interact most profitably with individual devices. So it's 521 00:44:16,159 --> 00:44:23,599 you know, it's all about sort of the big picture and interoperability. A 522 00:44:23,639 --> 00:44:28,239 lot of our episodes on the on the podcast here are focused on you know, 523 00:44:28,440 --> 00:44:31,960 program or logic controllers deep into heavily protected networks that nobody on the Internet 524 00:44:31,960 --> 00:44:36,920 has any hope of having a look at. If we're out on the internet, 525 00:44:36,960 --> 00:44:42,880 if we're reporting power usage, you know, the when we're out on 526 00:44:42,920 --> 00:44:49,320 the Internet, privacy is often a much bigger factor than it is deep into 527 00:44:49,360 --> 00:44:53,320 a heavily protected power plant network, you know. And so I've been asking 528 00:44:53,360 --> 00:44:58,960 you sort of questions from the perspective of the industrial security aspect. But you 529 00:44:59,000 --> 00:45:02,000 know, if my power or usages on the Internet, I do care about 530 00:45:02,159 --> 00:45:07,159 the you know, who can see that? Have you got privacy stuff built 531 00:45:07,199 --> 00:45:12,000 into this as well? And that's that's a great question because one of the 532 00:45:12,039 --> 00:45:15,519 first things I started out looking at security and cryptographic key management is all about 533 00:45:15,519 --> 00:45:19,840 security, is what I thought. Uh. And then when we started working 534 00:45:19,880 --> 00:45:23,360 with the State of Colorado and looking at several different programs, they had over 535 00:45:23,400 --> 00:45:28,559 seventy one programs last I heard a list of them that were candidates for this 536 00:45:28,679 --> 00:45:31,880 type of these type of technologies. Each one had different requirements, and I 537 00:45:31,920 --> 00:45:36,360 started asking, well, who makes a decision on them? And nobody raised 538 00:45:36,360 --> 00:45:39,159 their hand. There's nobody making a decision. When we started analyzing those their 539 00:45:39,199 --> 00:45:44,599 privacy questions there where is it held versus where's the store? What parts of 540 00:45:44,599 --> 00:45:46,960 it are private, what parts are public? Who has opt in is it 541 00:45:47,000 --> 00:45:51,840 opt out? Who who's allowed to have access, who's allowed to audit versus 542 00:45:51,840 --> 00:45:54,280 who is allowed to see it? Are you allowed to audit without seeing it, 543 00:45:54,400 --> 00:45:57,920 or they have to you get to know that they audit it. And 544 00:45:58,119 --> 00:46:02,639 there's so many legal and privacy questions, and that actually drove me to start 545 00:46:02,679 --> 00:46:07,880 asking these questions. We ran a campaign called Privacy for the People literally like 546 00:46:07,880 --> 00:46:12,000 four or five years ago. We got some international kind of attention and now 547 00:46:12,039 --> 00:46:15,159 I chair the Digital Privacy Initiative for I TRIPLEE, where we're looking at an 548 00:46:15,199 --> 00:46:20,000 international level to set up a lot of the boundaries to answer those questions. 549 00:46:20,360 --> 00:46:22,360 But things that we had to develop in the technology from the ground up is 550 00:46:22,400 --> 00:46:28,599 like every operation, every transaction has a public and a private categorization, and 551 00:46:28,639 --> 00:46:30,840 you tag the data whether it goes on the chain or whether it has to 552 00:46:30,880 --> 00:46:36,320 be held privately offline, and then what to do with it becomes the question 553 00:46:36,519 --> 00:46:40,320 after that or as different parts of that and what is required. We went 554 00:46:40,360 --> 00:46:46,880 through the pandemic learning that certain things are socially we overpower and override, and 555 00:46:46,920 --> 00:46:50,519 you have to have these things public so that we can all find out who 556 00:46:50,559 --> 00:46:52,639 has a disease and the numbers and everything, But what are your rights in 557 00:46:52,679 --> 00:46:59,199 that balance? And those are kind of the privacy questions and led to a 558 00:46:59,280 --> 00:47:04,280 whole government. It's architecture with forty two dimensions of governance that you can find 559 00:47:04,320 --> 00:47:07,079 in my dissertation if you pull that up. But this will be applied many 560 00:47:07,119 --> 00:47:14,760 places because it really mimics the real world how we manage our governments as addressed 561 00:47:14,800 --> 00:47:21,039 into technology applications and putting it into operation. So also a very exciting piece 562 00:47:21,599 --> 00:47:23,880 of what I've done, and I think more of what I will give to 563 00:47:23,960 --> 00:47:29,880 the world is along that lines than even the technology piece, because that's where 564 00:47:30,239 --> 00:47:35,199 really I think we're making a difference in giving people those rights in the digital 565 00:47:35,239 --> 00:47:39,599 era. This has been great. Thank you for these insights. It's a 566 00:47:39,679 --> 00:47:45,000 more complicated space than I realized before we let you go. Can you sum 567 00:47:45,079 --> 00:47:49,159 up for us what should we take away from here? What should we be 568 00:47:49,159 --> 00:47:52,400 thinking about? So it's time to get involved. Right now we are in 569 00:47:52,440 --> 00:47:58,360 the process of tokenization. Do an initial token offering for the distributed ledger. 570 00:47:58,760 --> 00:48:04,159 Look for it, come in. It's not an investment, it's a a 571 00:48:05,280 --> 00:48:08,880 a utility token. Sorry forget, but it is a way to get in 572 00:48:09,000 --> 00:48:13,920 early and more. Offering some discount rates as well. On the cryptographic key 573 00:48:14,039 --> 00:48:21,480 management talk to reach out to block frame tech dot com and we're bringing in 574 00:48:21,920 --> 00:48:27,280 vendors right now to integrate it into their product to design this next generation security 575 00:48:27,599 --> 00:48:30,000 today and have it in your next product release and when it comes down the 576 00:48:30,079 --> 00:48:35,039 road. It's like I keep saying, I'm excited about it, and I 577 00:48:35,079 --> 00:48:37,719 think we're changing the world. But I'll let you make that decision. But 578 00:48:37,760 --> 00:48:45,119 come talk to us. Andrew. That concludes your interview with doctor Gorrog. 579 00:48:45,280 --> 00:48:50,480 Do you have any final summary about what we talked about here to leave us 580 00:48:50,480 --> 00:48:53,719 off? Yeah, I mean, you know, to me, it's all 581 00:48:53,760 --> 00:49:00,639 about device networks, you know, backing away just a moment. The biggest 582 00:49:00,800 --> 00:49:04,480 denial of service attack in history, if you recall what a year ago or 583 00:49:04,480 --> 00:49:10,119 so two years ago was because internet connected household cameras had defects that were exploited 584 00:49:10,119 --> 00:49:14,159 and all of them attacked, you know, one one or two sites on 585 00:49:14,159 --> 00:49:16,440 the Internet at the same time. Not an example of trust. That's an 586 00:49:16,480 --> 00:49:21,920 example of just scale. The scale we're talking about is, you know, 587 00:49:22,159 --> 00:49:24,000 the number of these devices out in the world are just getting more and more, 588 00:49:24,000 --> 00:49:29,079 and a lot of them are Internet connected, you know, and when 589 00:49:29,119 --> 00:49:34,639 we connect an incredible number of devices across the Internet there's privacy issues, there's 590 00:49:34,760 --> 00:49:38,519 you know, there's verification issues. Are these power readings that I'm getting from 591 00:49:38,599 --> 00:49:44,360 rooftop solda producers in my geography? Are these are these you know readings real? 592 00:49:44,440 --> 00:49:47,280 Should I really pay these people? Are these traffic sensors from the neighboring 593 00:49:47,360 --> 00:49:50,960 city that say, I've got a problem coming my way, there's a traffic 594 00:49:51,039 --> 00:49:54,079 jam coming my way, you know, take corrective action? Are these real? 595 00:49:54,280 --> 00:49:59,199 You know, we're talking about about trust, We're talking about scale, 596 00:49:59,800 --> 00:50:04,800 you know, the unprecedented scale. We're talking about interoperability between vendors, between 597 00:50:04,880 --> 00:50:09,119 vendors and utilities, between utilities and other utilities, you know, probably even 598 00:50:09,159 --> 00:50:14,480 other applications that I just don't get yet. So yeah, it's uh, 599 00:50:15,480 --> 00:50:17,719 it looks like a kind of technology that we're going to see more and more 600 00:50:17,760 --> 00:50:23,880 of on the Internet and you know, with industrial and other kinds of applications 601 00:50:23,920 --> 00:50:30,400 going forward. So it's a space that I've tried to ignore, and I 602 00:50:30,480 --> 00:50:34,519 don't know that I can anymore. Okay, Well, thank you to doctor 603 00:50:34,559 --> 00:50:39,360 for doctor Okay, well, thank you to doctor Christopher Gorgh for speaking with 604 00:50:39,440 --> 00:50:43,639 you. Andrew, and Andrew, as always, thank you for speaking with 605 00:50:43,679 --> 00:50:46,199 me. It's always a pleasure. Thank you Nate. This has been the 606 00:50:46,280 --> 00:50:52,599 Industrial Security podcast from Waterfall. Thanks to everyone out there listening