1 00:00:03,920 --> 00:00:07,160 Welcome to the Everyday PM podcast, the podcast where we 2 00:00:07,160 --> 00:00:09,840 discuss project management principles for your everyday 3 00:00:09,840 --> 00:00:12,320 life. My name is Anne Campia, I am the 4 00:00:12,320 --> 00:00:16,280 host and founder of the Everyday PM, and today we're diving into 5 00:00:16,280 --> 00:00:18,920 one of the most transformative and could potentially be a 6 00:00:18,920 --> 00:00:21,720 little bit controversial topics in construction project 7 00:00:21,720 --> 00:00:25,120 management. It's the intersection of AI, 8 00:00:25,200 --> 00:00:29,040 artificial intelligence, ethics, and the future of planning and 9 00:00:29,040 --> 00:00:32,000 scheduling. Our guest today is Doctor Mala, 10 00:00:32,000 --> 00:00:35,880 who is a Senior PM professional with over 17 years of global 11 00:00:35,880 --> 00:00:39,320 experience leading planning, scheduling, and project controls 12 00:00:39,320 --> 00:00:43,920 for complex infrastructure programs exceeding $47 billion. 13 00:00:44,160 --> 00:00:48,040 So Doctor Malla is joining us for a series of these podcast 14 00:00:48,040 --> 00:00:51,440 episodes where we are going to really dive into various topics 15 00:00:51,440 --> 00:00:53,120 on construction project management. 16 00:00:53,480 --> 00:00:56,800 If you have not been introduced to Doctor Malla in a previous 17 00:00:56,800 --> 00:00:59,840 episode that you've listened to Doctor Malla, why don't you take 18 00:00:59,840 --> 00:01:02,160 a second to introduce yourself to our audience? 19 00:01:02,600 --> 00:01:05,440 Yeah. Thank you so much and for this 20 00:01:05,640 --> 00:01:09,120 wonderful topic that we are going to have conversation with 21 00:01:09,560 --> 00:01:13,480 and giving a very brief overview about myself. 22 00:01:13,880 --> 00:01:17,640 I've been in the construction landscape for over more than 17 23 00:01:17,640 --> 00:01:21,960 years as Anne has mentioned. And with respect to the kind of 24 00:01:21,960 --> 00:01:24,920 diverse projects experience, especially in the project 25 00:01:24,920 --> 00:01:29,600 controls roles starting from the scheduling, planning and the way 26 00:01:29,600 --> 00:01:32,360 sorts of contracts management operations. 27 00:01:33,160 --> 00:01:37,440 I bring in the wealth of different perspective lenses. 28 00:01:37,440 --> 00:01:44,200 And along with this, the tools with which the importance of the 29 00:01:44,200 --> 00:01:48,280 various control systems that are necessary for projects 30 00:01:48,560 --> 00:01:52,640 especially the mega scale and mega infrastructure projects is 31 00:01:52,640 --> 00:01:56,720 something like which I have gained over a period of time. 32 00:01:56,720 --> 00:01:59,960 And the tools, especially with respect to technological 33 00:02:00,000 --> 00:02:04,680 advancements, which is going at a meteoric pace is something 34 00:02:04,680 --> 00:02:12,680 which we need to engage not just ourselves, but also the peers 35 00:02:12,680 --> 00:02:16,600 and colleagues in different levels of maturity with which 36 00:02:16,600 --> 00:02:22,880 they want to have to get in the AI related realm to their 37 00:02:23,120 --> 00:02:28,480 specific domains. So apart from this, I would like 38 00:02:28,480 --> 00:02:34,320 to also add that I'm not just only focused on the practical 39 00:02:34,320 --> 00:02:37,000 aspects of the industry related experience. 40 00:02:37,560 --> 00:02:42,880 I also have gained the research acumen through my doctoral 41 00:02:42,880 --> 00:02:46,080 studies. So it's a it's a combination of 42 00:02:46,120 --> 00:02:51,920 scientific way of analysing, analysing the various types of 43 00:02:52,160 --> 00:02:54,200 projects that I have accomplished. 44 00:02:55,480 --> 00:02:58,040 Yeah. And it's it's a combination of 45 00:02:58,040 --> 00:03:03,160 research theory plus practical implication is something which 46 00:03:03,160 --> 00:03:09,040 I'm honoured to have got over these 17 years. 47 00:03:09,880 --> 00:03:15,200 So without Much Ado, if you want to know my profile a little bit 48 00:03:15,280 --> 00:03:21,920 of more extensively, I hope you would be in a position to replay 49 00:03:21,920 --> 00:03:25,680 the Episode 1, which was discussed on construction 50 00:03:25,680 --> 00:03:31,600 workforce necessity in the US industry and US, USA 51 00:03:31,600 --> 00:03:34,760 construction industry, specifically the mega projects. 52 00:03:35,200 --> 00:03:38,560 So I have explained a bit more about it. 53 00:03:38,560 --> 00:03:43,240 So in today's episode, we will have some sort of conversations 54 00:03:43,720 --> 00:03:47,080 with the kind of AI exposure that I have put into the 55 00:03:47,080 --> 00:03:52,960 practical applications and how does it, what works and what 56 00:03:52,960 --> 00:03:55,320 didn't work for me is something like which I would like to 57 00:03:56,040 --> 00:03:58,520 discuss about. Yeah, absolutely. 58 00:03:58,520 --> 00:04:02,240 And you know, the thought of AI, especially in construction 59 00:04:02,240 --> 00:04:05,720 project management, you know, as we start to integrate AI more 60 00:04:05,720 --> 00:04:09,480 and more into our project workflows and we see the 61 00:04:09,480 --> 00:04:13,640 predictive analytics that come from AI, there's questions that 62 00:04:13,640 --> 00:04:16,040 project managers are facing now, right? 63 00:04:16,079 --> 00:04:18,839 Like, how do we ensure the tools are used responsibly? 64 00:04:19,279 --> 00:04:22,920 Who's accountable if AI makes a mistake, right? 65 00:04:22,920 --> 00:04:25,440 We can't just assume AI is correct every time. 66 00:04:25,800 --> 00:04:29,400 How do we balance efficiency with ethical considerations? 67 00:04:29,680 --> 00:04:32,520 I think all of those are things that project managers are 68 00:04:32,520 --> 00:04:35,440 curious about, especially in the realm of construction project 69 00:04:35,440 --> 00:04:37,360 management. So I'm so excited, Doctor Malla, 70 00:04:37,360 --> 00:04:40,480 that you're here to bring that expertise as well as the 71 00:04:40,480 --> 00:04:46,200 academic side of it, plus the practicality of using AI with 72 00:04:46,240 --> 00:04:50,240 project workflows. So why don't we dive into what I 73 00:04:50,240 --> 00:04:53,920 wanted to ask you first? So you had mentioned previous in 74 00:04:53,920 --> 00:04:57,040 the first episode of this series as well as today that you've 75 00:04:57,040 --> 00:05:00,880 worked on some very, very large construction projects, very 76 00:05:00,880 --> 00:05:06,080 complex, large budgets. So you've published research on 77 00:05:06,160 --> 00:05:10,240 AI applications in the AEC industry as well. 78 00:05:10,600 --> 00:05:13,720 So let's start with a quick reality check, Doctor Mala. 79 00:05:14,000 --> 00:05:17,320 Where is AI actually making a difference in construction 80 00:05:17,320 --> 00:05:21,520 planning and scheduling right now versus where it's still just 81 00:05:21,520 --> 00:05:24,760 this kind of hype or this myth around AI? 82 00:05:25,120 --> 00:05:28,560 And can you share a specific example from your work where AI 83 00:05:28,560 --> 00:05:31,080 has genuinely improved project outcomes? 84 00:05:32,440 --> 00:05:37,000 Well, that's a really a very great place to begin with, 85 00:05:37,320 --> 00:05:44,560 because the very moment when a person mentions of AI at least 86 00:05:44,680 --> 00:05:48,440 for five years ago for I mean like pre COVID, at least in the 87 00:05:48,440 --> 00:05:53,080 construction industry, it was a kind of a buzz. 88 00:05:53,480 --> 00:05:58,240 It wasn't that kind of a buzzword or it was a mere 89 00:05:58,240 --> 00:06:04,520 thought like AI couldn't penetrate or get get assimilated 90 00:06:04,520 --> 00:06:08,080 into the construction industry. Oh, it's just the IT folks. 91 00:06:08,080 --> 00:06:11,160 It's not the construction industry professionals who who 92 00:06:11,160 --> 00:06:14,000 needs to learn? That was a kind of notion pre 93 00:06:14,000 --> 00:06:17,400 COVID. It was there and it so happened 94 00:06:18,240 --> 00:06:23,720 at a skyrocketing pace, the hype and the kind of implementation 95 00:06:23,720 --> 00:06:28,080 with which AI is happening in different domains, not just in 96 00:06:28,080 --> 00:06:32,440 planning and scheduling, but in the different life cycles of the 97 00:06:32,440 --> 00:06:35,680 construction projects, right, Starting from the conceptual 98 00:06:35,680 --> 00:06:39,600 stage, initiation, planning, execution, monitoring and 99 00:06:39,600 --> 00:06:41,760 controlling, closing and even maintenance. 100 00:06:42,040 --> 00:06:51,640 You have the kind of use cases built to this in, in in just a 101 00:06:51,640 --> 00:06:54,560 span of five years. I mean, at least there is a 102 00:06:54,560 --> 00:06:59,600 surge of interest that has been shown vehemently by various 103 00:06:59,600 --> 00:07:05,440 construction professionals and coming to the purview of 104 00:07:05,440 --> 00:07:08,960 planning and scheduling systems. I would bring like from my 105 00:07:08,960 --> 00:07:12,360 experience both as a practitioner as well as a 106 00:07:12,360 --> 00:07:18,320 researcher, I'd say that AI is not revolutionising the 107 00:07:18,320 --> 00:07:24,040 construction per SE, but it is meaningfully improving the 108 00:07:24,280 --> 00:07:28,760 various parts of the planning, components, coordination and 109 00:07:28,760 --> 00:07:33,600 decision support where the fundamentals are absolutely 110 00:07:33,600 --> 00:07:41,440 intact and it has got good data. So having data is not the right 111 00:07:42,360 --> 00:07:47,120 way with which AI can be implemented, but having the good 112 00:07:47,200 --> 00:07:50,720 data, data points, data structuring, information 113 00:07:50,720 --> 00:07:55,640 structuring is something which is key to implement any sort of 114 00:07:57,640 --> 00:08:00,640 AI related models, something like that. 115 00:08:01,120 --> 00:08:07,760 So if I look at where the real value is being delivered in the 116 00:08:07,760 --> 00:08:12,240 various projects that have been recently involved, it's 117 00:08:12,520 --> 00:08:17,560 primarily in the areas of building information modelling 118 00:08:18,040 --> 00:08:22,040 based virtual prototyping at the early project stages. 119 00:08:23,160 --> 00:08:27,960 So this is more of like 4 dimensional BIM. 120 00:08:27,960 --> 00:08:33,520 So a combination of the project schedule with the 121 00:08:33,520 --> 00:08:36,919 three-dimensional model and trying to integrate the various 122 00:08:36,919 --> 00:08:40,159 information components and trying to look at the sequence 123 00:08:40,360 --> 00:08:44,480 sequencing and at the conceptual level how it's going to get 124 00:08:44,480 --> 00:08:47,640 built is something like can be visualised. 125 00:08:47,640 --> 00:08:53,080 So that's the first point. The second one is like key is 126 00:08:53,080 --> 00:08:56,880 the information structuring and ensuring proper interface 127 00:08:56,880 --> 00:09:01,000 management is established in the projects. 128 00:09:01,760 --> 00:09:07,520 And the second, the second most important aspect, construction 129 00:09:07,520 --> 00:09:12,320 industry, especially the previous projects, everyday they 130 00:09:12,560 --> 00:09:17,160 produce humongous amounts of data in the form of daily 131 00:09:17,160 --> 00:09:21,360 project report, lots of information there and lots of 132 00:09:21,480 --> 00:09:26,680 progress site photographs on these sort of information that's 133 00:09:26,680 --> 00:09:31,440 being generated in the project site at daily basis is something 134 00:09:31,440 --> 00:09:36,560 where this information whether it is cleaned properly. 135 00:09:36,560 --> 00:09:40,680 So who is going to verify the reports? 136 00:09:41,000 --> 00:09:44,840 I mean like up until now majority of the construction 137 00:09:44,840 --> 00:09:48,160 projects were like maintaining this sort of information or 138 00:09:48,160 --> 00:09:52,960 database more of like auditing, more of like mere compliance 139 00:09:52,960 --> 00:09:54,800 kind of system. Yeah. 140 00:09:55,680 --> 00:10:00,920 So there was no sort of compliance of the data quality 141 00:10:00,920 --> 00:10:08,120 that happened. But with the kind of surge in AI 142 00:10:08,120 --> 00:10:14,040 related approaches that we want to implement, it's necessary 143 00:10:14,040 --> 00:10:17,280 that data quality is of immense importance. 144 00:10:18,680 --> 00:10:22,840 So that forms the basis with which the various algorithms 145 00:10:22,840 --> 00:10:26,720 that we can apply to the data that has been generated from the 146 00:10:26,720 --> 00:10:29,680 projects. So the third most important 147 00:10:29,680 --> 00:10:36,280 component which I can think it has delivered with respect to AI 148 00:10:36,280 --> 00:10:41,480 was identifying the patterns in the schedule activities with 149 00:10:41,480 --> 00:10:45,560 respect to risks and any sorts of trends in the historical 150 00:10:45,560 --> 00:10:48,520 project data. Although this can be done 151 00:10:48,520 --> 00:10:52,760 performed even on Excel, Microsoft Excel and other tools 152 00:10:52,760 --> 00:10:58,960 that are commercial tools that are available with the help of 153 00:10:59,080 --> 00:11:05,400 AI, different types of algorithms using different deep 154 00:11:05,400 --> 00:11:08,680 learning techniques or natural language processing techniques 155 00:11:08,680 --> 00:11:13,280 and all we can perform lots of. We can dissect the data and 156 00:11:13,280 --> 00:11:17,840 visualize it in different manner so that the visualization and 157 00:11:17,840 --> 00:11:22,480 scenario testing is something which can be performed aptly 158 00:11:22,480 --> 00:11:25,720 rather than just going with autonomous decision making. 159 00:11:25,880 --> 00:11:28,960 And this is where the 4th point that I would like to highlight 160 00:11:28,960 --> 00:11:34,040 how it has impacted and got better deliverables. 161 00:11:34,320 --> 00:11:40,120 So the first one is BIM based virtual prototyping in the 162 00:11:40,120 --> 00:11:44,400 initial stages of the project. Second one is the information 163 00:11:44,400 --> 00:11:47,240 restructuring and managing of the interfaces. 164 00:11:47,840 --> 00:11:51,960 Third one is it's able to give us various sorts of anomalies 165 00:11:51,960 --> 00:11:53,480 through the pattern recognitions. 166 00:11:54,040 --> 00:11:56,880 And the 4th and final one is like the visualisation and 167 00:11:56,880 --> 00:12:02,040 scenario based testing that if if we have humongous data, I 168 00:12:02,040 --> 00:12:05,560 mean like it's, it's quite interesting that all these years 169 00:12:05,720 --> 00:12:08,640 majority of the construction projects were maintaining 170 00:12:08,680 --> 00:12:13,440 databases which weren't of good quality. 171 00:12:13,680 --> 00:12:19,400 I mean, data cleaning and all was not that of utmost 172 00:12:19,400 --> 00:12:24,600 importance or sort of accuracy needs to be maintained, 173 00:12:24,680 --> 00:12:27,480 maintained is is not that kind of a notion that was prevailing. 174 00:12:28,160 --> 00:12:32,680 But if you want to implement AI systems and it's necessary that 175 00:12:33,680 --> 00:12:39,520 data needs to be of good quality and that's where the key point 176 00:12:39,520 --> 00:12:42,440 is. So with respect to my research 177 00:12:42,440 --> 00:12:48,520 arena, I would like to discuss whether it's a agile based BIM 178 00:12:48,880 --> 00:12:52,400 or the building information modelling or whether it is lean 179 00:12:52,400 --> 00:12:57,160 agile integration which I have performed research. 180 00:12:57,440 --> 00:13:03,560 It had shown significant improvements in the outcomes, 181 00:13:04,040 --> 00:13:10,240 which is not AI related stuff because that happened purely 182 00:13:10,480 --> 00:13:15,920 owing to the data discipline integrating the process based 183 00:13:15,920 --> 00:13:19,160 systems and more clarity of the information flows. 184 00:13:19,880 --> 00:13:24,800 So for instance, I can give an example like one of the water 185 00:13:24,800 --> 00:13:28,560 infrastructure projects that I have worked six years ago, the 186 00:13:28,560 --> 00:13:34,280 BIM models that were used for virtual prototypes combined with 187 00:13:34,280 --> 00:13:41,360 the structural various sorts of schedules or the planning logic 188 00:13:41,800 --> 00:13:48,040 had significantly reduced downstream level rework or the 189 00:13:48,320 --> 00:13:53,160 coordination failures because we are in a position to visualize 190 00:13:53,160 --> 00:13:56,400 it prior to its breaking down on the ground. 191 00:13:56,960 --> 00:14:02,520 So there wasn't a sort of neural network or AI kind of stuff 192 00:14:02,760 --> 00:14:06,960 being involved. However, this virtual try 193 00:14:07,280 --> 00:14:12,880 virtual prototyping definitely had helped in decision making. 194 00:14:13,720 --> 00:14:17,960 There was intelligence that has been embedded in how the 195 00:14:17,960 --> 00:14:23,800 information was created, validated and shared and where I 196 00:14:24,280 --> 00:14:30,000 Whereas how I see the hype where AI is positioned is as a 197 00:14:30,000 --> 00:14:32,360 substitute for a project understanding. 198 00:14:32,880 --> 00:14:37,840 So bringing in AI is not trying to bring in a human. 199 00:14:37,920 --> 00:14:42,040 So you cannot substitute humans judgement Right? 200 00:14:42,160 --> 00:14:46,720 So no algorithm is going to compensate a human judgement. 201 00:14:46,960 --> 00:14:49,920 Compensate for poorly defined scope yeah. 202 00:14:50,440 --> 00:14:55,360 Or weak contract management or immature project governance. 203 00:14:55,800 --> 00:15:01,920 So AI can be considered as a force multiplier or an amplifier 204 00:15:02,520 --> 00:15:09,600 to to to your good systems, but it cannot definitely take away 205 00:15:09,600 --> 00:15:14,000 the jobs. So AI is already valuable as a 206 00:15:14,000 --> 00:15:19,400 decision supporting tool over here, but only it sits on top of 207 00:15:19,480 --> 00:15:22,520 strong processes that are governing, which is the 208 00:15:22,520 --> 00:15:26,880 information management. Especially you have the various 209 00:15:26,880 --> 00:15:31,760 sorts of BIM standards for information management. 210 00:15:32,600 --> 00:15:36,880 Then there are also lean thinking principles and human 211 00:15:36,880 --> 00:15:42,640 judgment. Definitely adds keep role while 212 00:15:42,640 --> 00:15:47,000 implementing AI systems. So this is what my take on and. 213 00:15:47,480 --> 00:15:51,240 Yeah, yeah, absolutely. I think it's, I, I think in a 214 00:15:51,240 --> 00:15:54,520 nutshell, what we're learning is it's AI is available, it's 215 00:15:54,520 --> 00:15:58,240 there, but it's not completely there to take over our jobs. 216 00:15:58,240 --> 00:16:01,040 I kind of think we mentioned that even in Episode 1 as we 217 00:16:01,040 --> 00:16:04,440 were talking about that topic as well. 218 00:16:04,440 --> 00:16:09,000 So I'm, I'm curious as we dive a little bit deeper, because we 219 00:16:09,000 --> 00:16:13,080 need to address the elephant in the room, which is who's 220 00:16:13,080 --> 00:16:16,160 responsible when things go wrong, especially as you're 221 00:16:16,160 --> 00:16:21,520 working with AI systems to, let's say do project planning, 222 00:16:21,520 --> 00:16:25,680 project scheduling, making decisions, analyzing delays, 223 00:16:27,080 --> 00:16:30,080 predicting project risk decisions that can affect 224 00:16:30,080 --> 00:16:32,720 billions of dollars and thousands of jobs, right? 225 00:16:32,720 --> 00:16:35,880 Especially on some of the projects that you've had the 226 00:16:35,880 --> 00:16:39,680 opportunity to work on. So I know you've done extensive 227 00:16:39,680 --> 00:16:42,800 work in forensic schedule analysis and delay claims. 228 00:16:43,080 --> 00:16:47,160 So how does AI change the accountability landscape, and 229 00:16:47,160 --> 00:16:50,240 what are some of the ethical frameworks project managers 230 00:16:50,240 --> 00:16:52,280 should be considering or thinking about? 231 00:16:53,200 --> 00:16:57,800 Yeah, Yeah. This is really a very close. 232 00:16:58,920 --> 00:17:03,480 This is a kind of an interesting and debatable question also. 233 00:17:03,680 --> 00:17:07,319 Yeah. So at 1:00, on one hand, 234 00:17:08,720 --> 00:17:16,359 majority of the organisations are craving for embracing AI 235 00:17:16,359 --> 00:17:20,839 into their systems. So if you don't implement AI, it 236 00:17:20,839 --> 00:17:26,839 seems like many of their counterparts would be of the 237 00:17:26,839 --> 00:17:31,600 belief that we are outdated. Yeah, that's definitely the the 238 00:17:31,640 --> 00:17:33,320 sentiment we're all feeling right? 239 00:17:33,320 --> 00:17:35,880 Yeah, on. The other hand, there's another 240 00:17:35,880 --> 00:17:43,640 notion that whoever is I mean that's trying to promote or 241 00:17:44,000 --> 00:17:50,720 embed AI is also at the cusp of the integrity of the data 242 00:17:51,080 --> 00:17:55,840 project data, then non disclosure clauses, NDA's that 243 00:17:55,840 --> 00:17:59,440 has been signed by the various project professionals in the 244 00:17:59,440 --> 00:18:06,400 construction industry and and what sort of ethical validity 245 00:18:06,440 --> 00:18:11,080 are we working with like what sort of ethics with which are we 246 00:18:11,080 --> 00:18:14,720 working with when we use these AI systems? 247 00:18:15,880 --> 00:18:20,000 Definitely accountability is something like which we give to 248 00:18:20,560 --> 00:18:27,760 a person, right. So when you try to utilize AI, 249 00:18:28,440 --> 00:18:34,040 we cannot put a blame stating that this decision has been 250 00:18:34,040 --> 00:18:36,440 given by AI, so AI needs to be accountable. 251 00:18:38,280 --> 00:18:43,200 OK, then where to which Penal Code do we have to account to 252 00:18:43,200 --> 00:18:47,200 that person? Look it, it's so weird to get to 253 00:18:47,200 --> 00:18:53,840 this point, however lot of us have this sort of notion. 254 00:18:53,840 --> 00:19:00,120 So I would like to throw the kind of the kind of experience 255 00:19:00,120 --> 00:19:03,920 and exposure that I have while I'm use utilizing AI. 256 00:19:04,520 --> 00:19:08,480 So this is definitely a very question which is very close to 257 00:19:08,480 --> 00:19:12,560 my domain of experience, especially the forensics delay 258 00:19:12,560 --> 00:19:17,880 analysis which is completely dealing with claims, disputes, 259 00:19:18,280 --> 00:19:24,960 litigation and various sorts of contractual obligations. 260 00:19:25,240 --> 00:19:29,640 So in this construction industry, decisions are not just 261 00:19:29,640 --> 00:19:37,520 simply quite ad hoc or abstract. Any decision that's taken is 262 00:19:37,560 --> 00:19:43,240 affecting the money, safety, various sorts of reputation of 263 00:19:43,560 --> 00:19:49,840 the organization and livelihoods of the various employees who are 264 00:19:49,840 --> 00:19:54,000 living on on this industry. So accountability cannot be 265 00:19:54,040 --> 00:19:58,360 outsourced or being thrown to an algorithm which is developed by 266 00:19:58,360 --> 00:20:02,640 an AI or something like that. One of the dangers I see with 267 00:20:02,640 --> 00:20:09,400 the growing tendency of how AI systems is being AI outputs AI 268 00:20:09,400 --> 00:20:13,760 systems being utilized is the objective truth. 269 00:20:15,320 --> 00:20:18,760 You give certain information context to an AI system or a 270 00:20:18,760 --> 00:20:23,120 large language model and it brings out to you deliverables 271 00:20:23,120 --> 00:20:27,080 or responses. So it's it's a person's 272 00:20:27,240 --> 00:20:33,960 judgement to analyse the output with which the LLM has tried to 273 00:20:33,960 --> 00:20:37,320 give needs to be validated. Sure. 274 00:20:37,680 --> 00:20:43,200 It's if it is that depends on the kind of decision that's 275 00:20:43,200 --> 00:20:49,600 being made. So I don't feel that depending 276 00:20:49,600 --> 00:20:54,560 upon, I mean like correcting minor documentation or 277 00:20:54,560 --> 00:20:57,200 grammatical or kind of proofreading kind of stuff is 278 00:20:57,200 --> 00:21:04,040 something like, OK, but then every operation that you perform 279 00:21:04,040 --> 00:21:11,320 on AI algorithms or systems, it needs to have a kind of 280 00:21:11,320 --> 00:21:15,120 proofreading, a kind of validation. 281 00:21:15,360 --> 00:21:21,360 Sure. And human judgement cannot just 282 00:21:21,480 --> 00:21:27,160 go with AI. AI say whatever it says, but AI 283 00:21:27,160 --> 00:21:32,960 systems are only as neutral as the data assumptions or 284 00:21:32,960 --> 00:21:37,240 contractual context, whatever that you give or training it on. 285 00:21:37,760 --> 00:21:42,600 If the historical data itself is providing some sort of biased 286 00:21:42,760 --> 00:21:47,080 practices or adversial contracts, contracts or 287 00:21:47,080 --> 00:21:53,520 practices or incomplete records, then definitely AI cannot 288 00:21:53,520 --> 00:21:56,480 identify whether the data is correct or not. 289 00:21:56,800 --> 00:21:59,320 Right. The data, I mean whatever you 290 00:21:59,320 --> 00:22:02,840 feed and it tries to operate and then give you the output. 291 00:22:03,320 --> 00:22:07,560 But how would you know whether the data that you fed is the 292 00:22:07,560 --> 00:22:10,400 right one? And if you feel it's the right 293 00:22:10,400 --> 00:22:15,760 one, also, how would you justify the output or response with 294 00:22:15,760 --> 00:22:18,720 which it has given that you can rely on? 295 00:22:19,120 --> 00:22:22,880 So it's something on the experience, part of your 296 00:22:22,880 --> 00:22:25,920 experience, part of your domain knowledge. 297 00:22:25,920 --> 00:22:31,040 Expertise plays a key role and AI just plays a kind of an 298 00:22:31,040 --> 00:22:38,280 assistant to you to subvert some of your time consuming tasks or 299 00:22:38,440 --> 00:22:43,720 the tasks which would take level of effort in documentation or 300 00:22:43,840 --> 00:22:47,040 providing various sorts of standard procedures or standard 301 00:22:47,040 --> 00:22:51,360 operating practices. It's somewhere you can try to 302 00:22:51,480 --> 00:22:55,520 utilize certain AI components. And with respect to 303 00:22:55,800 --> 00:22:59,640 accountability standpoint, AI doesn't change the fundamental 304 00:22:59,640 --> 00:23:02,440 principles. Humans are finally responsible 305 00:23:02,440 --> 00:23:08,480 for it and what does AI change is how responsibility must be 306 00:23:08,480 --> 00:23:11,720 structured. AI can help you in structuring 307 00:23:11,720 --> 00:23:15,560 your responsibilities between the team. 308 00:23:15,920 --> 00:23:23,120 So it gives you multiple options so that you can try to minimize 309 00:23:23,200 --> 00:23:28,720 the time spent in organizing the stuff or trying to structure 310 00:23:28,720 --> 00:23:32,320 some mechanism. So it's good to have AI to 311 00:23:32,320 --> 00:23:34,800 brainstorm. So suppose if you don't have 312 00:23:34,960 --> 00:23:39,640 particular team members that you want to main brainstorm, so you 313 00:23:39,640 --> 00:23:43,880 can brainstorm your ideas and probably you might get some sort 314 00:23:43,880 --> 00:23:50,760 of leads or some sort of different sort of hybrid, hybrid 315 00:23:50,760 --> 00:23:54,960 analysis that you wanted to do. Probably that's the what if 316 00:23:54,960 --> 00:23:58,240 scenario analysis platform, which you can do all sorts of 317 00:23:58,240 --> 00:24:02,840 your whatever that you're thinking, try to utilize it as a 318 00:24:02,840 --> 00:24:05,000 platform to do the scenario analysis. 319 00:24:05,400 --> 00:24:09,960 And based on my research into this BIM contracts interface 320 00:24:09,960 --> 00:24:15,040 management or the dispute resolution, some of the some of 321 00:24:15,040 --> 00:24:18,400 the book reviews as well as one of the papers that have been 322 00:24:18,400 --> 00:24:22,200 involved, a couple of papers. I understand that the 323 00:24:22,200 --> 00:24:25,880 organizations need to clearly define who is owning the AI 324 00:24:25,880 --> 00:24:29,960 generated insights. So a lot of governance needs to 325 00:24:29,960 --> 00:24:32,720 be developed. Yeah. 326 00:24:32,720 --> 00:24:37,240 Otherwise it's like people go to the rabbit hole. 327 00:24:37,880 --> 00:24:41,600 Yeah. Establishing proper human in the 328 00:24:41,600 --> 00:24:44,640 loop decision protocols, although you're utilizing 329 00:24:44,640 --> 00:24:47,840 certain AI tools like for instance, in the forensic 330 00:24:47,840 --> 00:24:52,800 schedule delay analysis. While we while we prepare a 331 00:24:52,800 --> 00:24:56,680 report, final report on the various sorts of scheduled delay 332 00:24:56,680 --> 00:24:58,960 analysis that we take into the approach. 333 00:24:59,960 --> 00:25:05,120 It's the the most time consuming part is providing that report. 334 00:25:06,040 --> 00:25:10,720 So probably in those scenarios, structuring the report, writing 335 00:25:10,720 --> 00:25:15,960 the clear chronological narrative of the events, trying 336 00:25:15,960 --> 00:25:22,480 to transcribe the various sorts of narratives with respect to 337 00:25:22,480 --> 00:25:27,760 delays given by the site in charge of superintendents, it 338 00:25:27,760 --> 00:25:33,160 can bring out the teams, bring out certain causalities what 339 00:25:33,160 --> 00:25:37,640 caused these delays. So you can utilize it as a tool 340 00:25:37,680 --> 00:25:43,840 and utilizing in such kind of analysis and that that's the 341 00:25:43,840 --> 00:25:51,720 part which is consuming a lot of level of effort from the final 342 00:25:51,720 --> 00:25:54,680 decision rather than the final decision point. 343 00:25:55,120 --> 00:26:01,560 So when the AI systems can be utilized in these sorts of 344 00:26:02,160 --> 00:26:08,640 minute tasks, it would definitely help the schedule 345 00:26:09,160 --> 00:26:13,680 delay experts or for forensic schedule and quantum delay 346 00:26:14,000 --> 00:26:18,480 professionals in spending majority of the time in bringing 347 00:26:18,480 --> 00:26:22,800 out the analytical component into picture and trying to 348 00:26:22,800 --> 00:26:27,320 analyse the substantiation of the cost with explainable and 349 00:26:27,320 --> 00:26:32,280 auditable components with the help of AI outputs that they 350 00:26:32,280 --> 00:26:36,280 get. And aligning the AI usage with 351 00:26:36,280 --> 00:26:40,520 the contractual and legal frameworks is also another 352 00:26:40,520 --> 00:26:43,880 important aspect. So in this forensic schedule 353 00:26:43,880 --> 00:26:49,080 delay analysis only suppose there are various contractual 354 00:26:49,120 --> 00:26:52,720 entitlement for delays of different types, excusable, non 355 00:26:52,720 --> 00:26:56,760 excusable, various types of concurrent pacing delays, all 356 00:26:56,760 --> 00:27:00,200 these sorts of contractual obligations that is provided, 357 00:27:00,920 --> 00:27:05,960 which is like quite voluminous. It's, it's really helpful when 358 00:27:06,000 --> 00:27:10,280 AI systems are used to try to interpret the clause. 359 00:27:11,160 --> 00:27:15,800 Not everybody is an legal expert, right? 360 00:27:15,800 --> 00:27:22,040 Like especially the freshers who wants to join or have a flavour 361 00:27:22,040 --> 00:27:24,640 of this sorts of scheduled delay analysis. 362 00:27:24,640 --> 00:27:29,600 I think working with AI in trying to understand the various 363 00:27:29,600 --> 00:27:32,960 contractual clauses before it is being entitled to a particular 364 00:27:32,960 --> 00:27:38,160 delay is definitely helpful. So when it's simple like when 365 00:27:38,160 --> 00:27:42,040 you have humongous data, AI would definitely help you in 366 00:27:42,040 --> 00:27:45,800 different use cases. But identifying that different 367 00:27:45,800 --> 00:27:48,920 use case is up to the human to know. 368 00:27:49,560 --> 00:27:54,640 And that can come when that particular professional is adept 369 00:27:55,040 --> 00:27:59,200 or expertise in their particular domain and has got vast 370 00:27:59,200 --> 00:28:02,480 knowledge on the projects that they have worked. 371 00:28:04,240 --> 00:28:07,560 Viewing from different perspectives is something like 372 00:28:08,080 --> 00:28:10,880 it. It gets developed when you have 373 00:28:11,040 --> 00:28:14,040 an exposure on or an experience. Yeah. 374 00:28:14,040 --> 00:28:18,360 So in claims and disputes especially, AI should never be 375 00:28:18,360 --> 00:28:22,640 the final authority. At best it can help you in 376 00:28:22,640 --> 00:28:26,480 bringing out those patterns, themes or inconsistencies. 377 00:28:27,240 --> 00:28:33,200 But causality, I mean getting the entitlement responsibility. 378 00:28:33,200 --> 00:28:36,480 But responsibility, who is going to be responsible for particular 379 00:28:36,480 --> 00:28:39,760 delay, whether it is a contractor, owner, subconsultant 380 00:28:40,120 --> 00:28:44,920 consultant is primarily determined by the professionals 381 00:28:44,920 --> 00:28:49,760 who understand the context and contracts and the case law. 382 00:28:50,320 --> 00:28:55,440 And especially it's not AI's who are working on the project, it's 383 00:28:55,560 --> 00:28:57,360 the construction professionals, right? 384 00:28:57,360 --> 00:28:59,200 They're they're working with the projects. 385 00:28:59,560 --> 00:29:02,200 So they should be final accountable people. 386 00:29:02,200 --> 00:29:05,560 It's not definitely. AI, not AI. 387 00:29:06,320 --> 00:29:09,160 And finally, I would like to add this ethically. 388 00:29:09,520 --> 00:29:15,840 Project managers need to ask themselves that do not ask just 389 00:29:15,840 --> 00:29:22,000 can we use AI, but they should be thinking like should we and 390 00:29:22,240 --> 00:29:27,120 under what governance should be working with this AI systems. 391 00:29:27,280 --> 00:29:29,880 Yeah. So definitely organizations need 392 00:29:29,880 --> 00:29:35,280 to design ethics into their systems right from day one and 393 00:29:35,280 --> 00:29:42,160 not just in not just doing a retrofitting kind of assessment 394 00:29:42,160 --> 00:29:48,520 or just when things go wrong. So before launching AI and a 395 00:29:48,520 --> 00:29:53,920 sort of understanding the gap, understanding the various sorts 396 00:29:53,920 --> 00:29:59,640 of information management standards, processes, workflows 397 00:30:00,720 --> 00:30:05,920 and pilot projects are necessary before launching that that that 398 00:30:05,920 --> 00:30:10,920 could be a kind of what if scenario, yeah. 399 00:30:11,040 --> 00:30:16,440 Or or a kind of scenario testing possibility that could be done, 400 00:30:16,680 --> 00:30:19,040 yeah. Absolutely, absolutely. 401 00:30:19,040 --> 00:30:21,760 Well, Doctor Malla, that was incredibly insightful. 402 00:30:22,480 --> 00:30:26,640 I think you've given us at least the listeners of, of our podcast 403 00:30:26,640 --> 00:30:29,640 episodes of balanced perspective that couples really well with 404 00:30:29,640 --> 00:30:33,640 what we covered in episode 1 of this series around building 405 00:30:33,640 --> 00:30:36,680 skills from the ground up. I think there continues to be a 406 00:30:36,680 --> 00:30:39,160 theme and what you're sharing in your research as well as what 407 00:30:39,160 --> 00:30:42,360 you're seeing out on the field, which is the human element and 408 00:30:42,360 --> 00:30:46,600 how that can really partner well with AI, but not necessarily 409 00:30:46,600 --> 00:30:50,480 acknowledge AI as you know, another stakeholder in the room 410 00:30:50,480 --> 00:30:54,400 that we can hold accountable to if if something goes wrong. 411 00:30:54,400 --> 00:30:59,440 So I appreciate the balanced perspective you acknowledging AI 412 00:30:59,440 --> 00:31:02,720 is real potential while being clear eyed about all the ethical 413 00:31:02,720 --> 00:31:05,600 challenges that we still need to continue to remember to address 414 00:31:05,600 --> 00:31:08,560 as project professionals. So some of the key takeaways 415 00:31:08,560 --> 00:31:11,800 from this episode that I would share with the audience start 416 00:31:11,800 --> 00:31:15,880 practical, as Doctor Mala Point pointed out, focus on some of 417 00:31:15,880 --> 00:31:19,440 the AI applications that are working today rather than 418 00:31:19,440 --> 00:31:23,360 chasing what the hype of the future of AI is, is, is going to 419 00:31:23,360 --> 00:31:26,760 be build those accountability frameworks. 420 00:31:26,760 --> 00:31:29,920 So before even thinking about implementing AI, you should be 421 00:31:29,920 --> 00:31:32,960 thinking about the decision making protocols that governance 422 00:31:32,960 --> 00:31:37,160 structure, that Doctor Mal pointed out, audit your training 423 00:31:37,160 --> 00:31:41,000 data and involve diversity in the stakeholders that are 424 00:31:41,000 --> 00:31:46,280 looking at the AI system design and then augment, don't replace. 425 00:31:46,280 --> 00:31:49,840 I think for me, that was a very strong message and what Doctor 426 00:31:49,840 --> 00:31:53,200 Molla shared today, which is focusing on AI as a tool that 427 00:31:53,200 --> 00:31:56,200 enhances human judgement rather than replaces it. 428 00:31:56,200 --> 00:32:00,280 So that was again, incredibly insightful and balance for what 429 00:32:00,280 --> 00:32:03,080 could have potentially be a very controversial topic. 430 00:32:03,080 --> 00:32:05,720 So Doctor Molla, thank you for bringing that expertise to the 431 00:32:05,720 --> 00:32:08,840 podcast today. For folks that want to continue 432 00:32:08,840 --> 00:32:11,360 the conversation with you, where can they find you online? 433 00:32:12,880 --> 00:32:16,920 Well, you I would appreciate Anne if you could share my 434 00:32:17,440 --> 00:32:20,960 LinkedIn. So if you could just type on my 435 00:32:20,960 --> 00:32:23,600 name. So definitely you would find me 436 00:32:23,600 --> 00:32:28,720 and we can can get connected. And I would highly appreciate 437 00:32:29,320 --> 00:32:34,840 Anne and the Everyday PM podcast for enabling me to share some of 438 00:32:34,840 --> 00:32:40,760 the best practices, knowledge, exposure, so it's more of like 439 00:32:40,760 --> 00:32:44,680 conversational kind of topics. And I truly enjoy. 440 00:32:45,600 --> 00:32:49,560 And anyone who wants to connect with me and learn more about any 441 00:32:49,560 --> 00:32:52,040 sort of collaboration, yes, I'm open to it. 442 00:32:52,800 --> 00:32:56,320 It's been, yeah, truly an honor and a pleasure hosting you here. 443 00:32:56,560 --> 00:32:59,520 I'm very excited to release these again as a series. 444 00:32:59,520 --> 00:33:02,800 As Doctor Mal had mentioned, a good conversation to listen to 445 00:33:02,800 --> 00:33:05,000 from start to finish in these episodes. 446 00:33:05,000 --> 00:33:06,720 So thank you so much for joining us. 447 00:33:06,720 --> 00:33:09,360 If you'd like to continue the conversation with me, you can 448 00:33:09,360 --> 00:33:11,880 find me on LinkedIn as well. I'll drop that link into the 449 00:33:12,360 --> 00:33:15,120 podcast description. Make sure to follow and 450 00:33:15,120 --> 00:33:17,720 subscribe to the Everyday PM Podcast. 451 00:33:17,720 --> 00:33:20,320 You can find it on every podcasting platform. 452 00:33:20,640 --> 00:33:23,160 And let us know what you thought about today's episode. 453 00:33:23,160 --> 00:33:25,920 So thank you so much for joining us today and for the important 454 00:33:25,920 --> 00:33:29,160 work you're doing to ensure AI serves the construction industry 455 00:33:29,160 --> 00:33:31,400 responsibly. Dr. Mala, first and foremost, 456 00:33:31,400 --> 00:33:33,640 thank you for that. And to our listeners, as you 457 00:33:33,640 --> 00:33:37,920 explore AI in your own projects, remember technology is a tool, 458 00:33:37,920 --> 00:33:41,080 but ethics is a choice, so make it intentional. 459 00:33:41,440 --> 00:33:43,320 And thanks for listening to our episode. 460 00:33:43,320 --> 00:33:45,720 And until next time, take care.