How well will a robot function as the source of marketing communication? Advertisers spend huge sums to recruit just the right (human) endorser for a brand, but at least so far no one seems to be giving much thought to what a salesbot or AI-generated model should look like or sound like.
That’s a big mistake. To read more, please click here.
These days, seemingly everyone is applying Artificial Intelligence (AI) and machine learning. I have written about disruptions in the manufacturing industry, such as Industry 4.0, while illustrating the Hard Trends that indicate where improvements will be made in the future.
The construction industry, which makes up 7% of the global workforce, should already have applied these technologies to improve productivity and revolutionize the industry. However, it has actually progressed quite slowly.
Growth in the construction industry has only been 1% over a few decades while manufacturing is growing at a rate of 3.6%. With the total worker output in construction at a standstill, it is no surprise that the areas where machine learning and AI could improve such statistics were minimal. Yet, those technologies are finally starting to emerge in the industry.
Artificial Intelligence (AI) is when a computer mimics specific attributes of human cognitive function, while machine learning gives the computer the ability to learn from data, as opposed to being specifically programmed by a human. Here are ten ways that AI and machine learning will transform the construction and engineering industries into what we’ll call “smart construction.”
Even efficient construction teams are plagued by cost overruns on larger-scale projects. AI can utilize machine learning to better schedule realistic timelines from the start, learning from data such as project or contract type, and implement elements of real-time training in order to enhance skills and improve team leadership.
When a building is constructed, the sequence of architectural, engineering, mechanical, electrical, and plumbing tasks must be accounted for in order to prevent these specific teams from stepping out of sequence or clashing. Generative design is accomplished through a process called “building information modeling.” Construction companies can utilize generative design to plot out alternative designs and processes, preventing rework.
The construction process involves risk, including quality and safety risks. AI machine learning programs process large amounts of data, including the size of the project, to identify the size of each risk and help the project team pay closer attention to bigger risk factors.
A recent startup utilized 3D scanning, AI and neural networks to scan a project site and determine the progress of specific sub-projects in order to prevent late and over-budget work. This approach allowed management to jump in and solve problems before they got out of control. Similarly, “reinforcement learning” (machine learning based on trial and error) can help to collate small issues and improve the preparation phase of project planning.
Professionals often fear machines will replace them. While intelligent machines will take over first repetitive and eventually more cognitively complex positions, this does not mean a lack of jobs for people. Instead, workers will transition to new, more fulfilling and highly productive roles to save time and stay on budget, and AI will monitor human productivity on job sites to provide real-time guidance on improving each operation.
Manual labor not only has the potential to be taxing on the body, but also to be incredibly dangerous. Presently, a general contractor is developing an algorithm that analyzes safety hazards seen in imagery taken from a job site, making it possible to hold safety briefings to eliminate elevated danger and improve overall safety on construction sites.
AI and machine learning have the capacity to plot out accurate distribution of labor and machinery across different job sites, again preventing budget overruns. One evaluation might reveal where a construction site has adequate coverage while another reveals where it is short staffed, thereby allowing for an efficient and cost-effective repositioning of workers.
When structures can be partially assembled off-site and then completed on-site, construction goes faster. The concept of using advanced robots and AI to accomplish this remote assembly is new. Assembly line production of something like a wall can be completed while the human workforce focuses on the finish work.
The data gathered from construction sites and the digital lessons learned by AI and advanced machines are all tools for improving the productivity of the next project. In this way, each construction site can contribute to a virtual textbook of information helpful to the entire industry.
Structures are always settling and shifting slightly. It would be beneficial to be able to dive back into data collated by a computer to track in real time the changes and potential problems faced by a structure — and AI and machine learning make this possible.
Given the inevitable changes on the horizon, and the potential for costs to drop up to 20% or more with increased productivity, professionals in the construction industry must pay attention to Hard Trends, become more anticipatory, and ultimately learn to turn disruption and change into opportunity and advantage.
Discover proven strategies to accelerate innovation with my latest book The Anticipatory Organization. Follow this link for a special offer.
Shape the Future–Before Someone Else Does It For You!
Plenty has been written about how AI gets smarter with experience, the way people do. If you perform a task 100 times, you are faster and better at that task the 100th time than the first time–and AI models have that same quality. The more experience they have (usually expressed as the more data they have seen), the more patterns the models can recognize to make better sense of each new thing they see. I do a lot of work with AI models around website customer experience–often focused on how web users search and navigate company websites. The AI models reveal insights of where web users get stuck, or, more happily, which content seems to answer their needs.
That’s very powerful, but even more powerful is connecting AI models to automated actions. You see, if all the models do is to provide better insight to humans, those models are useful, but they will always be gated by the time and cost of humans taking actions on those insights. I have heard clients express to me in frustration that “the last thing [they] need is another dashboard”–even a smarter one populated with keen AI insights, because it still leaves them with more and more manual improvements to make.
My recent work with SoloSegment [full disclosure: I am a Senior Strategist and partner with SoloSegment] has opened my eyes to how AI can lead to immediate and continuous improvement of a website. You can use behavioral data to make searches on the site more successful. You can recommend content based on what has worked for others in the past. In other words, your website becomes more autonomous–a living, self-improving entity–that gets better the more people use it.
None of this means that you don’t need people to do the vast majority of the tasks of creating content, improving design, and all the rest of the things we do for our websites. But, for the first time, there are some things that humans don’t need to do, because the AI models, coupled with automated actions, can make some of the improvements in hands-free fashion. I don’t know about you, but this feels like a breakthrough to me, where we finally have linked the intelligence of the models to quickly and automatically improving the customer’s experience. And I can’t help but think there is much more to come.
I am working with large companies on their use of Artificial Intelligence all the time, and it is possibly the most polarizing technology I have ever been involved with. Some people believe that AI will give us all a life of leisure, with machines doing more of the work so we don’t need to slave away for 40 hours a week. Others are spooked because they think that AI is coming for our jobs. What seems hard for each of those groups is that they are both essentially projecting the same thing–it’s just a question of whether they are optimistic or pessimistic personalities.
I see the same thing with my own clients–AI is equally polarizing, but this time it is around its effectiveness. Some are AI skeptics, talking about how the technology is over-hyped. Others believe it is magic, and will buy anything with those two magic letters. Both views are right–and wrong. AI just isn’t very simple.
Businesses should always be looking to improve their return on investment, which means choosing the simplest technique that solves the problem. Sometimes that’s AI, but often it’s something simpler, cheaper, and lower risk, so we should start there. Many folks are surprised when I say that, because they expect me to be pushing AI for everything, but I don’t see how that makes any sense. I spoke with a potential new client who was so taken aback that as we were leaving, they said to us, “Gee, we speak to a lot of vendors, but thanks for surprising us.”
If you are listening to vendors blathering on about that 5G blockchain kind of AI, it’s time to stop listening to buzzwords and start looking for competence. If it sounds too good to be true, it is. If your spidey sense starts to tingle every time they start talking about neural networks, listen to that inner voice. AI is no different from every other kind of approach out there. Used appropriately, it can be a huge benefit to your business. But you should be asking questions if your vendors wave their hands and can’t really explain why AI is needed and exactly why it works better. Don’t pay surge pricing for the flavor of the month.
I have spent the last 40 years on the cusp of various technologies. (It’s a trick. If you are on the cutting edge, there are no experts, so you get to call yourself one.) Now I am an expert in Marketing and AI. (See what I did there?)
I actually have been working in text analytics since the 80s and was first exposed to machine learning in IBM Research in the 90s, so I have been doing this for a while, if that counts for anything. So I am used to hearing people talk about how AI is the future. And it is.
Sometimes, it’s just how you talk about it. I remember early in my career, I did what I thought was a knockout presentation on some new superpower technology, and as the audience was filing out, a few people came up to speak to me afterwards. They were all very excited and all agreed as one person breathlessly said to me, “Wow, you are really a visionary.”
Except that’s bad. Because that means that they didn’t think they needed to do anything about that technology for three years. So if every time you hear about machine learning it sounds to you like Big Data 5G Blockchain, then you are missing the power of the present.
Machine learning can take the data you are sitting on and start predicting outcomes that you needed to wait to have happen. We are working with clients to predict the bounce rates of new pages without having to wait three months to see what they are. You can imagine applying the same approach to exit rate, social shares, inbound links, and any other content metric.
Think about what an advantage that is. Rather than suffering with poorly-performing pages for months until the data stabilizes, you can make changes presuming that those pages will perform the way similar pages have in the past. So make them look like better-performing pages instead. But do it now, not months from now.
That is what machine learning does. It takes all the data that you already have and speeds up the correct decision. That speed is your competitive advantage. Or at least it is your competitive advantage if you are using machine learning now. Conversely, if you think AI is the future, then it might be your competitor’s advantage now.
