About this episode

This episode of the Coffee With Digital Trailblazers was sponsored by Quickbase.

Speakers

• Isaac Sacolick – President of StarCIO. Bestselling author. 1,000+ articles on StarCIO, CIO.com, and InfoWorld. Keynote speaker. CIO and Digital Transformation influencer. Thinkers360 top IT, DevOps, and Agile Leader.
• Tim Douglas – GM of Construction at Quickbase
• Joanne Friedman – CEO/Principal of Smart Manufacturing at Comnnektedminds Inc, Industry 4.0 and Digital Transformation Coach
• Joe Puglisi – Chairman, North Andover Investors. Veteran CIO in construction and manufacturing
• Heather May – Founder/President of May Executive Search.
• Martin Davis – Managing Partner, DUNELM Associates. Veteran CIO in manufacturing.
• Liz Martinez – MD at The PMO Whisperer

About the Coffee With Digital Trailblazers

The event is hosted by Isaac Sacolick on LinkedIn Fridays at 11 am ET. The event attacts digitial transformation leaders, from CXOs to team leaders, who learn from experts on driving change in their organizations. Every week we explore a topic and share lessons learned, and all are welcom to attend. Visit https://starcio.com/coffee/next-event which will redirect to the next event.

Recordings of the events are available to StarCIO Digital Trailblazer Community members .Review the community’s advisory, leadernng, and leadership programs.

Episode Transcript

Isaac Sacolick:

Get started and we’ll get going. Hello everyone. Welcome to our 103rd episode of the Coffee with Digital Trailblazers. I’m Isaac Sacolick. I want to welcome our normal guest, Joe and Joanne. I think Martin will be joining us today. I think John is missing this week, and I want to give an early shout out to our sponsor this week. Our sponsor is Quickbase. Quickbase is a work management platform used by more than 12,000 organizations, and I want to welcome Tim Douglas GM for construction over at Quickbase. We’re just going to give everybody a few minutes to get started before we get started here, and then we’ll get right into our topic today, which is AI era, pragmatic ways to transform work management in complex industries. And for those of you working in, I dunno, industry five, oh, construction, manufacturing, even government, healthcare, some aspects of financial services, we’re going to get into a conversation of pragmatic AI and pragmatic ways to transform work management.

And we’re going to discuss some of the nuances that what a complex industry is and how to navigate applying AI in those industries. So that’s going to be our conversation today. Totally excited to see all of you joining here today. Just going to give it a couple of more minutes before we get into our topic. Martin, you made it. Glad to see you on stage. We have a few people here who are experts. We have our Joanne and Martin, our Mark manufacturing experts. We have Joe who’s a construction and manufacturing expert, and I’m your transformation expert with some background in manufacturing and construction to share how we apply ai, machine learning automation in pragmatic ways in complex industries. So glad you’re all here for this episode today. I want to welcome Tim Douglas. Tim is a general manager over at for construction over at QuickBase. Again, QuickBase is our sponsor today, QuickBase is a work management platform used by more than 12,000 organizations worldwide. Tim, I want to start with you and let’s just start painting the picture of, just break down some terminology for everybody. What is work management and what makes an industry complex? Welcome, Tim.

Tim Douglas:

Thanks. Thanks so much for having me. Excited to be here. I think that when we look at work management, it’s how organizations really structure the execution of the work, the digital aspects, the physical aspects. And I think that as you talk about industry 5.0 in particular industries like manufacturing and construction, these are industries that rely significantly on the choreography of people, equipment, materials, things that happen in an office versus things that happen on a shop floor, things that happen in a trailer versus things that happen on a job site. And so there’s a lot of variability in the delivery of whether they’re delivering a product or a building. There’s a lot of things that need to be coordinated, including the data in order for them to be able to deliver their product or project on time with high quality and safety for the people doing the work.

Isaac Sacolick:

I think those are the intersection of physical and digital aspects, and that complexity of the choreography between ops and back office is a hallmark of what makes an industry complex. I also know that working in industries where safety is a factor, you mentioned safety and then quite frankly just low margins. And we’ve got to squeeze our profitability based on our operational excellence. And among things, I know the construction industry, what makes it complex for me is just the number of players involved between the owner, the general contractor, the architect, the subcontractors and trades, the equipment delivery, all the manufactured equipment that shows up. It’s just a complex choreography. I really like that word that goes into running a construction project. Let’s give the floor over to Joanne. Joanne, we were talking today about work management, what makes an industry complex. I’d love to hear your thoughts today on this.

Joanne Friedman:

Oh,

Isaac Sacolick:

Hi there. Dropped for a moment. Go ahead, Joanne.

Joanne Friedman:

Welcome back. Thank you. Sorry, you want to repeat the question for me

Isaac Sacolick:

Please? So we’re talking about complex industries and we’re defining work management for everybody here so that when we get into solutioning with pragmatic ai, we are grounded in what makes an industry complex and what work management is all about.

Joanne Friedman:

I think there’s two different aspects of it that I would highlight. The first is that complex industries does not necessarily mean that the product that’s being produced is complex. It’s more that the number of stakeholders in the value chain can be very large, where you have a lot of sort of co-operative situations where you have competitors that cooperate. So in manufacturing you have OEMs and outsource design and business process outsourcers and supply chain partners and other key stakeholders in various aspects of the production of a good, and it starts at a value chain situation where it could be an OEM that these days does not actually manufacture its own products. They are responsible for ideation and design, but then everything else is outsourced and they do the marketing and the branding and so forth. But really the processes of design, engineering, et cetera, are left to other people. So that would be a very complex industry. Think about high tech, think about automotive, any of the industries where hard goods are produced. And also in some cases things like pharmaceuticals or chemicals. These are all complex industries, so work management can be twofold. It can be on the production floor in one factory or it can be all of the work that it has to be managed across that big value chain.

Isaac Sacolick:

Got it. Joanne, I want to jump over to Martin, stay on manufacturing for a minute and then we’ll shift to construction with Joe,

Martin Davis:

Work management and manufacturer, Joan’s painted a big picture of how different companies interact and how can, so I’m going to maybe take it more to the manufacturing floor side of things. If you think about the complex manufacturing operation, you’ve got the interaction of a lot of different areas of work and you’ve got the business management side of things starting to interact with the operations management side of things and the actual operations of the business. So you’ve got all of those normal business type interactions. Then you’ve got all those manufacturing type interactions from the raw material moving into a warehouse, offering delivery, moving from a warehouse to a production floor, that material being consumed into the manufacturing and finished goods coming out the other end and all the different aspects of work that goes in and how that data that accompanies the work needs to transform, needs to move. That’s all part of work management. And if you think about the different aspects of how you support a manufacturing, so maintenance for example, you have a whole lot of maintenance work that goes on, how do you schedule that maintenance, how do you schedule downtime, all those different aspects. So there’s a lot of work management that surrounds and manufacturers not part of the finished goods as well. So all of those different pieces are all different aspects of work management.

Isaac Sacolick:

Got it. So again, back to that word, choreography of data, choreography of workflow of equipment. Joe, what’s the same and different in construction?

Joe Puglisi:

Well, Tim has already hit on I think one of the four or five key ingredients that make industries complicated. He’s talked about the multiple participants. I used to explain construction to people by likening it to making a movie. When you make a movie, you have so many participants, the people that write the scripts, the actors that execute the script, the cameramen, the lighting, the set design, and all of these people are either self-employed or belong to different and disparate companies. And coordinating in your terms, Isaac choreographing this to produce the film, which then goes through editing and final production and distribution, incredibly complex construction shares a lot of those characteristics in that you have lots of different companies and lots of different specialties that are involved. But let’s talk about some other things Joanne mentioned, supply chains and the logistics, that can be a complicating factor. I want to add to the list regulation and compliance. When we look at industries like construction, but more so finance, healthcare, some transportation industries, the regulations and the shifting of regulations can be very, very complicating in terms of managing the workflow. And then lastly, I’ll throw in the unpredictability of certain things that are beyond our control. In construction, one of the things that often hinders progress is the weather. You can’t always bore the concrete when it’s too cold. You can’t move materials, inclement weather. So unpredictable environmental types of issues can also be complicating factors.

Isaac Sacolick:

Thanks for bringing up unpredictability. I think that’s a good one. I want to go back to Tim. Tim, we came from this world where we believe projects in these areas we could put down together a nice clean looking Gantt chart and the scheduling would work almost as planned. And when you factor in all the different players, all the different stakeholders, as Joanne mentioned, the regulations that are involved, the unpredictability from weather, the coordination between back and front office, how does work management look different than that rigid Gantt chart that we picture in our mind?

Tim Douglas:

Yeah, it is definitely not rigid. That’s sure to be sure. And I think that the belief I’ve been in construction technology for over 20 years is the idea is that technology can snap all of these variables into line. And the fact of the matter is there’s no single system that will choreograph a supply chain or a job site. And that is one of the keys that I think has plagued the industry in their ability to drive efficiency gains, to drive major improvements in productivity, profitability because not only are choreographing different resources and people and those things, it’s the systems then that support them and the disconnected aspects of those systems that lead to a lot of this information falling through the cracks. And I think that that’s one of those areas that I still think that the most widely deployed tool in construction technology today is excel. And when you have information going from many different systems of record and the gaps are somehow being filled by uploads and downloads and Excel, it only exacerbates the problem. So I think that the belief is that technology can drive much more easy connections between all these parties, but the fact of the matter is as good as any technology is there still are gaps in the way it supports these workflows today.

Isaac Sacolick:

Go ahead, Joanne.

Joanne Friedman:

I definitely agree with Tim’s point, but I would also suggest that companies need to be forward thinking about the strategy that they employ around that data because you also have to incorporate, at least in many factories these days, those that are let’s say large businesses or large enterprise, the use of physical equipment like robots, cobots, and there are many, many moving parts and we not only have existing technical debt and silos around the data from our legacy systems, we now have new types of data that we have to integrate as well. So as things fall through those cracks, particularly in automation, robotics and cobots, you’re looking at a scenario where no matter how good the spreadsheet might be or the system of record might be the choreography, which is one of my favorite words by the way, as well as the orchestration have to be taken into account to accommodate for Joe’s point on regulatory requirement, think about companies that want to be sustainable. You have lots of silos, lots of interconnectivity, lots of interoperability challenges, and then you have to put that together in a way that not only is secure and compliant to regulation, but also flexible to meet the economic micro and macroeconomics of the industry.

Isaac Sacolick:

Look, what I picture when I think of what we’re talking about here is in the construction space, I’ve got CAD systems, I’ve got scheduling software, I’ve got multiple parties on the job site. I’ve got foreman and four people and superintendents who are trying to give me some feedback about where we’re operating on the job. I’ve got weather factors going on and just that picture of somebody working in the job trailer with a bunch of spreadsheets trying to make changes on the fly. Tim, I don’t think it cut its anymore. I think that’s what you were getting at, right?

Tim Douglas:

Yeah, no, exactly right. And I think what I really love that Joanne brought up is you have to start with the end in mind. And I think that that is one of the things that I think that we’ve seen in the application of technology is the things that these companies are trying to achieve in construction, profitability, safety, quality, those are universal, whether we’re talking about the biggest GCs in the world or the guy installing roofing on your house. And I think that what’s really interesting is when they start with the end in mind, when these companies think about what their objectives are and they try to get clarity around one of those things that are the big levers that drive success in their business, you have to not eat the entire elephant. You have to figure out where you’re going to start to get the greatest return in the shortest amount of time possible.

And I think too many people try to do too much too soon, the capacity of these organizations to absorb change, not just the adoption of the technology, but the change in the workflows that support it and the various data inputs is joined, brought up as well. There’s not just in the ERP or HRMS or project management systems in the office, but that information coming from the robots on the job site, the IOT information that’s being collected on a job site only gets more complex. And so that’s where going back to basics, the things that drive your performance as a business starting there really provides the best opportunity to solve a problem and then figure out what’s next.

Isaac Sacolick:

What I love about this conversation, Tim, is that I used to write the article for engineering news record back when I was working at Dodge about how construction and to some extent manufacturing was always lagging in its technology investments. And to some extent that’s still somewhat true, but I do think a lot of other industries that connect that physical and digital world where safety is a factor have a lot to learn from what construction and manufacturing are doing today. I mean, we’re still catching up on the technology side, but it’s that real time less than perfect plan that we always have to prepare for. That happens in hospitals, that happens in government, it’s happening in financial services now. And I think that’s what makes the interesting part of bringing our data together, bringing integration together and now looking for some pragmatic ways AI can be used and deliver business benefits.

I want to shift gears to that discussion. We’ve seen a lot of hype around AI for a lot of interesting reasons. We are also seeing Gartner, I think, put AI at the beginning down slope into the trough of disillusionment. So it’s at the peak of its hype cycle and I think we have a lot of people listening here who are looking for how do I get time to value? How do I get to realistic pragmatic use cases of AI that we can deliver some real tangible results? So Tim, I want to get this from everybody. Tim, we’ll start with you. What are some pragmatic AI business benefits and use cases?

Tim Douglas:

Yeah, and I think that again, it goes to focus. You don’t apply capital AI to your business. You really have to think about areas first and foremost where you have a reasonable and a reasonably consistent data set that you can leverage these models against to derive insights that are going to help you do your job better. And I think that comes to the personalization. Again, there definitely are advantages in looking at the data sets and the overall impact of the business that the executives love to gain those insights. But again, it’s not necessarily across all the aspects of the business. Maybe it’s driving more effective bid management and ensuring that you’re working with the right partners. Maybe it is driving safety because of the number of safety forms that you’ve digitized and have collected, gives you a dataset to drive insights around how you can more effectively drive safety. So again, it starts with focusing on a problem and not trying to apply AI across the business but solve a specific problem, and particularly for a specific person that makes them better, that gives them a copilot or a caddy to help them do their job better because it’s the accrual of those jobs being done better job after job that improves the models, which improves the insights which improves the return that AI can deliver.

Isaac Sacolick:

Tim, what are some examples that you’re seeing in construction?

Tim Douglas:

Yeah, I think one of the things that I’ve seen pop up recently is around safety and pre-task planning. Every action on a job site requires some type of pre-task plan to ensure that safety is adhered to in the execution of the work. So there are hundreds and hundreds of these forms. If these forms today are paper and done on a clipboard, your ability to derive any insight and have been in many job site trailers where you see the boxes of paper in the corner of all the pre-task plans, it’s really hard to derive insights from a box of paper. Companies are digitizing these relatively simple and straightforward pre-task plans, and then as they happen hundreds of times over and over again, you can begin to run these AI models against them to understand where issues aren’t being caught in the things that you need to do to drive a safer workplace on the job site.

Isaac Sacolick:

Excellent. Liz, welcome to the floor. Your thoughts on pragmatic AI use cases in complex industries.

Liz Martinez:

I love it. I love it. I’m listening to Tim and I love how you’re diving into going to where the pain points are, right? If you go to where the people are already having the pain, where are those stacks of paper? Where are those Excel spreadsheets? Identifying where those pain points are, you’ll actually find out where the business could be impacted the best. So AI often is confused with, oh, I’m going to be digitizing away people’s jobs and therefore people are afraid of it. But if you’re actually going to where people are working and they’re already having pain and not able to address some of the safety concerns, that’s a perfect place to start looking at where AI can make a difference and collecting that information. So there’s always two ways to do this, right? You got to be careful about garbage and garbage out, but if you have some good data that you could actually start training models on, it might be able to teach you something that you hadn’t seen before. So going top down on what data you actually already have to see what you can see is one way, but going bottom up and starting out with the people and seeing what data that you need to collect or what’s out there that could be digitized. I just love that approach. So thank you, Tim.

Isaac Sacolick:

Let’s go around the horn. Joe, I’ll tell you one of the use cases I saw in construction that excited me is just being able to validate that the person working in front of you from a particular subcontractor has all the training and certifications to be working in the capacity that they’re working in front of you. And it can do a real AI photo, a real AI lookup of their credentials and make sure they have all the certifications that are expected of that person doing that job. Pain points in construction or any solutions in construction that you’ve seen that are pragmatic to apply AI to

Joe Puglisi:

Going back 10 or 15 years, there’s a little technology called BIM, and this is, I dunno if you would consider it to be ai, but it’s certainly a precursor to, as Tim points out pooling digitized data to identify places where trades are going to collide. I think the term was collisions, collision detection. The sprinkler head was going to be installed right where there was a lighting fixture on the plans from the fire protection and lighting contractors. I love what Tim was saying earlier about using digital forms of data to find opportunities to reduce problems, reduce risk, and this leads to the motivation, and we haven’t talked about the motivation here. You have to change the culture from one in which every contractor is interested in delivering their product on the job. Tim, the lighting guy wants to put the lights up and the electrician wants to put the power in the building, but they don’t really care about each other. However, if you can identify spots where they will interfere with one another, and you can avoid those, they can each get their jobs done faster and avoid rework or delays. So finding the motivation I think is a great use and if we sprinkle a little more AI into the mix along with the fundamentals of bim, we can really make a difference.

Isaac Sacolick:

Let’s shift over to manufacturing for a second, Joanne. We’re looking for pragmatic ways to apply AI and manufacturing. I know you have a very long list. Let’s pick a couple of our,

Joanne Friedman:

Okay, first on the list, predictive maintenance, anything that was considered remote process automation in a manufacturing environment can easily become pragmatic AI because you’re already collecting the data and any corporation, big or small that is making a product. If you’re not collecting data as you’re going through it, even if it’s on a spreadsheet, you’re not going to be able to get the metrics that you need to determine whether you’re A making money off of the manufacturer of the product, or B, that the quality of the product is at par or better. Recently, we’ve seen several situations where we’ve had what I would call enforce errors. You had Ford MIS numbers, you had stellantis MIS numbers, you had a huge number of issues around food production that have been very newsworthy because people got sick and people actually died from quality. And I’m not trying to make light of that or be dismissive of it, that anybody would allow a factory to get into a position of making people sick. But that’s beside the point. That’s a personal opinion, but anything that’s a predictive analytic or a novelly detection root cause, these are things that can be done as programmatic changes and the use of AI makes ’em very pragmatic because they’re repetitive. So anything that was previously a repetitive process can be done as pragmatic ai, and I think that’s going to shift the landscape tremendously over the next year particularly because it’s going to leverage the complicated knowledge of the workforce.

Isaac Sacolick:

Thank you, Joanne. Martin, let’s get one example and then I will take my mid session break. Go ahead, Martin.

Martin Davis:

Joanne’s not the best one, predictive maintenance one, but I’ll go with another typical one is around downtime. If you’re collecting data about downtime and where you’ve had a problem online that’s caused the line to go down using AI to actually analyze that and understand common causes of downtime so you can tackle them, that makes your productivity, makes cost of goods less, your output higher. So there’s a quick one, there’s a thousand other ideas, but that’s a good one.

Isaac Sacolick:

Thank you, Martin. We’re going to shift gears after our break. We’ll start talking about more opportunities versus hype. I want to welcome everyone to our hundred third episode of the Coffee with Digital Trailblazers. We covered topics for those of you leading digital transformation in your companies. Today we’re talking about complex industries, very broad topic. We’re focused a little bit on manufacturing construction, but really any industry that’s connecting the physical and digital world has a lot to learn from this conversation. Today’s episode is brought to you by Quickbase as our sponsor. QuickBase is a work management platform used by more than 12,000 organizations worldwide to boost productivity when managing large scale projects and operations for industries like construction and manufacturing, bring together people data and process in one platform. Quickbase makes it easy to oversee and coordinate work across the business, minimizing administrative overhead, mitigating operational risk, and delivering project visibility to balance the demands of operational efficiency and costs while promoting employee safety.

For more information, please visit their website at htttps://www.quickbase.com. So folks, we’re going to talk about now shifting our gears to our listeners, our digital trailblazers, seeing all the hype out there, all the great articles that I like to write about that may be a little bit more farfetched and saying, how do we get to something that’s more realistic? How do we get to something that’s pragmatic? I love using Joanne’s time to value as a key KPI like going back to what Liz introduced and saying, let’s focus on a problem that really needs solving. I left everybody here a link to one of my articles that talked about a problem in the construction industry, and it’s simply talent. There is a talent gap in terms of bringing in more people into the industry. There’s a lot of people with a tremendous amount of knowledge that are aging out and being able to capture knowledge, make it easily accessible while you’re on the job site so you can find information about what you’re about to go install.

That’s a great use case. But also just demonstrating to employees that you’re a world-class organization, you’re efficient, you’re using technology in modernized ways, you’re making their lives safer and easier while they’re doing their work. Those are all things that I look for when I’m looking for AI use cases that can deliver tremendous business value. And so let’s go around the room, Tim, we’ll start with you again. We’re looking for how can digital trailblazers identify pragmatic AI opportunities versus hype, futuristic, or hard to execute ai? Thank you, Tim, and I’d love to hear your thoughts on this.

Tim Douglas:

Yeah, so first and foremost, I think it’s been said, but focus on the things that are common in core to your business, the things that have to do with people, resources, equipment in those buckets. Then start with the end in mind. What outcome do you want to drive across the insight that you’re seeking to gain by applying AI to those models? In the last segment I talked about pre-test planning, kind of a field-based workflow that has a potential given the high volumes of data that can be generated relatively quickly from were paper-based forms, but now could be digitized forms. I think it’s a great use case and it drives directly at driving more safety on the job site. I’d like to give another that maybe goes back to the office, and it is an area where customers have driven more standardized processes, even if they’re done in spreadsheets.

Today is in estimating and bid management companies have developed over the years very, very sophisticated datasets around their historic bids. And the performance of those is something that they don’t have insight into because the data is locked in 800,000 different spreadsheets for the bid packages for specific projects. But these teams do drive a lot of consistency in the way that they look at these bids. So it seems to me that there’s an area that is ripe for AI models to look into to better understand specialty contractor performance, material cost, product performance after the fact. So I think that that starting with a problem that can have a massive impact on the profitability of the project, on the backend offers, I think a really great way for folks to leverage those data sets, those spreadsheets to help them drive that more predictable, more profitable delivery of the job.

Isaac Sacolick:

Joe, you must love that because talk about where spreadsheets were being used completely and just trying to answer a question back to the CFO. Are we going to be able to deliver project X profitably based on bid y and is this a good area for the company to invest their resources and now we can start bringing that data in a consistent way and start doing some predictive analysis around that? What does that make you think of, Joe?

Joe Puglisi:

I love it, Jim. 800,000 spreadsheets in a bid package. I’m going to quote that one a lot. It’s great. It’s so true. In a low margin business where pennies can really make a difference using AI to go through past projects, performance of subcontractors, scheduling, snafus, all of this data that is contained within the history of projects could really lead to some incredible insights. And gosh, don’t you and I Isaac, wish we had that kind of power when we were on the line.

Isaac Sacolick:

I know I can’t tell you how many work in progress reports that I’ve seen in various forms and the data there is so valuable. But I want to switch over to Heather. Heather, I think you heard me say the of bringing talent into the construction industry. I’m sure you have something to share with us. For those who don’t know Heather, she’s an executive recruiter and is usually the person who’s going to give us some insights around talent. Go ahead, Heather.

Heather May:

You absolutely got my attention. Thank you. And I’m probably going to steal something that Joe would mention, and that’s about communication because as you do bring in younger people, I would communicate that they’re to them through the whole company, but to let them know that we’re thinking of new projects, new ideas, and we’ve talked about a hackathon, an AI hackathon, something that could be a competition, could be fun, could give the company a different perspective for younger employees and to be, and for the people that are working there, be ambassadors to get other people to want to look at the company as a prospective employer. But there’s a wealth of knowledge, yes, on 800,000 spreadsheets, but in the brains of people and the stuff that they talk about at coffee hour or over cocktails, it just makes them crazy, ’em try to fix it.

Isaac Sacolick:

Yeah. Heather, the other side of talent that I think about, Tim introduced the folks back in the office, the incoming workforce. They’re a lot more technically talented. They understand applications, they understand data a little bit. They’re not married to their spreadsheets. Like Joanne, our generation, we grew up with spreadsheets and we didn’t have the ability to create applications that fit the gray work that sits in between an ERP and a scheduling system and in a CAD system. And now these folks coming into the workforce, they want to do something and solve a problem and make sure it’s scalable and they need some guidance and tools to be able to do that. Go ahead, Liz.

Liz Martinez:

I was just thinking about how we actually call it actual intelligence. We were talking last night to a recruiter and she was saying how I have more in my brain than is on LinkedIn. I said, I call that actual intelligence, not ai. And then we moved into, I don’t know if you remember the swivel chair integration. You go from one system and then you turn around and go to another system using your brain as the interface. But anyway, back to the point, it’s interesting to think about how in the workplace that more and more, it’s not just what’s on the resume because the resumes can be tweaked to, as you’ve seen I’m sure, Heather in the workplace, that people can put whatever they want on resumes. People can lie, first of all, and it’s much easier to lie because you can tweak it with the AI tool to make it say whatever it needs to say, but the intangibles of what it takes to actually be a productive member of the team, it seems like that’s an open area that could really benefit from AI moving into the emotional intelligence, moving into the interactions among the team members.

And that seems like a very interesting area moving into the psychology of how teams perform and team performance. I mean, all of us know about Myers Briggs and disc and all those things, but that seems like an area of complexity that really could explode in the future. Joan, just a

Isaac Sacolick:

Thought. I love that. Liz, I want to go back to my question about digital trailblazers and pragmatic AI opportunities. Joanne, we did a video a number of months ago and you labeled yourself the pragmatic vision painter. So I want you to put that pragmatic vision painting hat on. Let’s just talk about how do we find, we talked about finding problem statements. We talked about data as a key element. We talked a little bit about focusing on people and using AI that provides impacts for people. What are some other ways that we could use AI in pragmatic ways?

Joanne Friedman:

Excuse me. Well, there are several. Let me move from manufacturing a product to, excuse me, the retail side of a product. If we were in a direct to consumer business or we were a major retailer, pragmatic AI is something that you can use to personalize a customer journey. You can really use AI in a way to target in and hone in, not on what sort of contrived marketing you want to give a consumer to buy a product, but you can really use it from an inventory perspective. You can use it to trace the customer’s journey to where to meet them best, not just where to meet them, meaning all channels like omnichannel, but which is the best channel to reach them and start looking at decomposing some of the huge amounts of data that are being produced on the consumer side to something that is far more manageable.

So we can look at the word pragmatic in many ways. What’s the outcome you want to achieve as an organization from digital transformation? What can you leverage a new set of tooling like AI beyond generative ai? I mean, if you look at all of the downside of the chatbot and took the information from the chatbot and actually read it through something that was more pragmatically oriented, you’d have far, far better results. So I mean, it is widely applicable and from the point of view of human in the workforce or human in the loop, which is I think what you were alluding to that tribal knowledge, inculcated, institutionalized knowledge is one of the largest overlooked assets in a corporation, and that comes directly from your workforce to Heather’s point. And it can be used to help them upskill, re-skill transition, all of those wonderful things.

Isaac Sacolick:

I think that tribal knowledge is a key issue. I mean, I think it’s a challenge in both construction and manufacturing because of the talent issue, but I think that extends to all organizations. When you think about being able to put agents up and put large language models up and say, I can put information at the tips of what people need to know when they need to know it, and that’s what’s happening on a job site, and that’s what’s happening on a plant. It’s real-time information against a huge corpus of information and the ability to provide that to them in a very, what used to be a very star trekky kind of feeling that we didn’t think we’d ever get to with this world of keyword searches and broken results that we can now fix and do natural language querying and put that directly embedded in our applications and make our workers a lot more safer and easier and more knowledgeable.

I’m going to share another one with you, Joanne. Here’s an even more rudimentary and pragmatic case. A joke probably attest to this. You’ve come up with your model, you come up with your data, and now all of a sudden you have an exception. You need to annotate something because something changed in your original model. There’s a weather factor, there’s a delay, there’s a new person joining the workforce who’s operating at 70% capacity. There’s all these real world scenarios that you’d like to track around your program, around your project, and guess what? Your ERP and your scheduling software doesn’t have that. And so what does the person do, Tim? They export the data, they create a new column in their spreadsheet to capture it, and now they’ve got derivative data that doesn’t go anywhere. So whenever you work on that problem, you have to annotate something that’s going to go into your model today into the future. This is that gray work area that we’re talking about where there’s a tremendous opportunity to build your applications out, build out your data science models out, and eventually use AI against them. Go ahead, Joanna. We’ll go back to Tim after.

Joanne Friedman:

Thanks. One of the things that I just wanted to point out is to exactly the situation you’re describing. I mean, root cause analysis is a big deal, but it could be root cause for any problem. It’s not just in manufacturing. It’s not just figuring out the scenario. But to your point about the scenarios in a factory, now you have, well, was the root cause the machine a person, the raw materials, the recipe, there’s so many variables and so many factors that go into that. This is one of the most difficult applications. But when you need to annotate something, whether it’s because your AI model drifted or your ML model drifted, or because of a new circumstance, think about this, and particularly from the point of view of a new digital trailblazer, data by itself has no value, but when you add context to that value, it becomes information and becomes very useful.

But context can change, and it’s the ability to constantly be and the flexibility to change the context to your point about anomalies or annotations, that makes it even more valuable. So if you’re careful in your thought processes and bring human beings into the equation as well, what you end up with is something very close to an expert system because you’re absorbing all of that context, giving it the flexibility and be choreographed as opposed to orchestrated to Tim’s point and giving it a capability that has a top line value and a bottom line value at the same time. So just by changing that context and being able to be flexible, you’re creating a capability for an organization that has levers from here until tomorrow, and also opportunity at the other side.

Isaac Sacolick:

Thanks, Joanne. Let’s go back to you, Tim. We’re talking about this gray area that creates the mess when we can’t manage things in rigid ways. Maybe we can dive into that a little bit with you.

Tim Douglas:

Yeah, I think one, go back to something that Heather brought up and something I’ve been talking about throughout is the role that people play in in doing their jobs. And I think particularly in construction, we are going through and it will only accelerate a massive generational shift in the workforce there, which is going to have a significant impact on that institutional knowledge base that has been developed and has evolved over the past 30, 40 years. And I think that this goes to how these companies think about how they are going to attract and retain talent while driving increased performance. And I think it comes back to how they establish themselves as thought leaders in the application of technology to drive improved ways of working. The kids coming out of college today did not come out of college to dive into 800,000 spreadsheets. They want to be a part of something.

They are very tech savvy. They want to make a difference, and their ability to do so in companies that rely on spreadsheets and disconnected systems and non-collaborative practices are not going to be very, very motivating for that next generation of worker in construction. And as we go through this workforce transition, I think this is going to be a critical element for any construction company to think about as part of their overall digital strategy. AI strategy. Again, it comes back to people and the next generation is going to have a major say in how that happens.

Isaac Sacolick:

Tim, I want to go down a question here. I just received from one of our listeners, and we are going a little bit backwards when we talk about 800,000 spreadsheets. Chris is going into the future. He says he’s curious how the panel sees additive manufacturing with things like 3D. Printing at scale is changing the game in terms of AI and work management. And I knew Joanne’s going to really want to step up around this one. It’s almost like we’re going to have a convergence, right? Between construction, manufacturing, we start bringing all that capability onto the construction floor. Go ahead, Joanne.

Joanne Friedman:

I’m trying to figure out how to break that down. There’s a lot to unpack in the question. I know Confluence. Yes. I think in terms of additive, I think in terms of things like the benefits that the aggregation will bring, think about photovoltaic resin, not to drop big words, but basically those are reflective of the sun. So while if it goes into a paint while you’re driving your car in the summertime, you’re absorbing sunlight that gets converted into energy. Voila, a way to charge your new car battery, that would be something that you would want to use AI for. And pragmatic ai, particularly because it would be by road constantly repeating the same process of how much is absorbed, how much is being converted, how much is now recharging your battery. But that would be one and the same photovoltaic resin, by the way, is now being used on windows in buildings. So that’s the convergence of the two industries.

Isaac Sacolick:

Anyone else have a thought around that on how 3D printing might be changing things? Martin, Joe,

Martin Davis:

I think we’re going to see a lot of use of 3D printing in specific circumstances. It really depends on the scale and speed and things like that. And as those change and improve, you’re going to see more. And the obvious one is spare parts in a manufacturing facility, do you hold a massive stock of spare parts or do you have the capability to photo print the parts you need? And so I think the scale and complexity of some of these things can be really good. The use of where Joan was going in terms of manufacturing and construction coming together. Previously you may be making wall panels on site. Then we went into the stage of, okay, you can make wall panels in a factory and just deliver the kit to sites, but now we’re getting into use of additive printing of buildings, so use of concrete as an additive material and actually doing it there. So I think you’re seeing these cycles of change happening of where different techniques are and speed of these techniques are growing and improving and allowing for new approaches to how you build a house and new approaches to how you do manufacturing.

Joe Puglisi:

Yeah, I think Isaac, that’s exactly right. The 3D printing, I think of it in a broader sense, the 3D printing of core and shell for buildings is really a large implementation of robotics and robotics and prebuilt components or componentization of construction is likely to be where the convergence occurs.

Isaac Sacolick:

Look, I think of this as in technology terms to some extent. We’re always finding ways to move up stack to do the combination of I’m going to automate more, collect more data, and then use artificial intelligence to help people make better decisions in real time. And what that’s requiring us to do is bring knowledge forward, bring capability forward to our people who are doing work today. The world is going to change very quickly over the next 10 years in terms of how we’re doing things and what that means. What gets me excited is as things get easier, we’re able to build more interesting capabilities into our buildings, into our plants, whether it’s making them more sustainable, whether it’s putting smarter iot devices in place, it’s that entire cycle of taking what was hard before, making it easier for people, making it faster, making it safer, and then we get to start working on some more interesting things. So I think I saw Joanne go off on mute, and then I’d like to hear from Tim one more time also around this,

Joanne Friedman:

Oh, I’m sorry, you faded out for a second. Isaac, can you please repeat what you were asking me?

Isaac Sacolick:

I’m asking your thoughts about how we’re moving up stack, how we’re constantly trying to solve for yesterday’s problems that were hard for us to do. We’re solving ’em today so that we can get at the more automation, more machine learning oriented and the more AI as the technology becomes more mainstream for us.

Joanne Friedman:

Well, I think we’re moving very quickly towards not generative ai, but regenerative AI or generalized ai. And I think we’re doing it in a way that is kind of step by step, but it’s happening much faster than people realize. And to that point, I would say as we’re moving up the stack, we have to be very careful about things like ethics. We have to be very careful about where that data is coming from that we’re using to solve those problems. But I think you’re going to start to hear a lot more about things like expert systems because it’s not just about a generative AI tool because you can only go so far with generative ai. Then you have to get into other kinds of ai, physics, ai, for example, or a specialized process or statistical process where you have to invoke those. But if the goal is to be solving problems higher and higher up the stack, then we also have to look more at how we define the word strategy and where words like competitive advantage and resiliency and agility actually come into play.

Isaac Sacolick:

Thank you, Joanne. Tim, let’s leave our listeners with some last minute thoughts from you about pragmatic AI today.

Tim Douglas:

Yeah, I worked for a leader many years ago who often said, the future is here. It’s just not evenly distributed. And that sticks with me, I think particularly in these days of how there’s so much out there that you could do. How do you begin to again, drive back to the focus, drive back to what’s pragmatic? And I think that when we talk about prefabrication or offsite construction, I think it’s a great example of the industry recognizing for certain scopes of work, what are we doing again and again out on a job site, cutting pipe, cutting metal studs, putting things together in a dusty, dirty weather impacted job site. When we can take these resources, put them in a clean facility more effectively, manage the flow of the people, the resources, the equipment, and then deliver bathrooms or head walls or whatever that thing is that they could drive those efficiencies, they started with the end in mind, how can we do this better?

And they worked backwards. They drove to a more integrated practice of what before had seemed like, how could we ever solve for this? We each have different competing agendas. And so I think as they integrate these practices, as they derive data sets around what happens as we move to a more prefabricated workflow, we’re going to be able to leverage AI to further enhance that. But a starts with people, starts with focus and starts with having an outcome-based end in mind approach to make these things real. And then driving the awareness in the industry that these things are working, they’re delivering real results. They’re not simply hype. And I think that that is accelerating. I think that that awareness in the industry is really accelerating and motivating other companies to think differently about how they’re working today, how they can compete today, how they can attract and retain talent today. And I think all of these things just really come together to help drive that transformation that we’ve been talking about in the industry for the last 25 years.

Isaac Sacolick:

Well, great conversation here. I’m hearing pragmatic AI is about finding your problem statement, aiming to deliver real results, focusing on problems to help people get us in a situation where we’re collecting better data, enabling more real time decisions, focusing in the physical and digital worlds, in our complex industries about making work safer, more enjoyable, more exciting for the people who are working there, capturing better knowledge so that we can bring in a younger workforce and a more pragmatic workforce ready to scale us to an industry or a set of industries now that are going to change considerably over the next decade. Fabulous conversation. I want to thank Liz, Joanne, Martin, Heather, Joe for joining us our special guest. Tim, thank you for being our general manager, our construction expert from our sponsor. Quickbase Quickbase work management platform gives complex industries the data visibility they need to keep large scale projects on time, in budget, and promote a culture of safety and compliance.

For more information, visit http://www.quickbase.com. I’d like to remind everybody our conversation next week. I’m just trying to find it on my list conversation next week. The 22nd will be about the boring use cases, the boring AI use cases, delivering significant business value. So we’re going to continue on with our conversation with ai. We’ll be taking a break on the 29th for the US Thanksgiving break, and then we’ll be back in December with more topics. Everybody. Thanks for joining. Thanks again to our sponsors, QuickBase. Thanks, Tim, Joe, Joanne, Liz, Martin, and Heather for joining me on stage. Those of you who are trying to find our next episode, please bookmark the url star cio.com/coffee/next event that will redirect you to all our upcoming events. Everybody have a safe, enjoyable weekend. I hope you found some pragmatic AI use cases to bring back to your companies. Thanks for all the applause here. Have a great weekend. We’ll see you here next week.