Cruising without a care in the world knowing that this drive, like every drive, will be interruption-free, with no traffic lights to stop, no unexpected congestion, simply you getting into your connected vehicle. Your car is not autonomous yet and the US infrastructure is still subpar to that which allows das autobahn to lift its speed limits, yet this future is easy for me to imagine because the tools are already here for the most part, it’s a matter of data collection (GPS data + destination) and allocation (access to modify traffic lights). Seeing the future vision is the necessary requisite for any project, this initial over-optimism creates the project champion, that will then be able to carry the concept through to fruition. That said, having worked for short of a decade leading innovation projects, one thing I’ve learned is it is always exponentially easier to promote change on the periphery than to act as an ancillary change agent to the day-to-day. This article though is about a journey and critique of construction tech, so I’ll start with this idea above of the seamlessly connected roadway then explain how it’s relevant to construction tech, and like every thought-provoking article will close with some open questions to the industry.The seamlessly connected roadway concept is simple, first, make it mandatory to enter your destination into the authorized application and then create the push and pull. The pull is the predictability of an optimized drive every time while the push is reworking traffic penalties for deviating from the prescribed route at the prescribed speed in the prescribed lane, etc. As I mentioned above, most of the tools are here, for example, Google Maps and a culture of traffic laws, the missing component is the political push and a master algorithm that could regulate traffic lights, bridges, trains, etc. to optimize the route in real-time. Obviously, this is an understatement to the complexity of the challenge to make something like this a reality, the point I’d like to highlight here is the importance of operating on the periphery when it comes to innovation. New unicorns are being minted as we speak for autonomous vehicles and/or electric car companies and the narrative almost goes unquestioned. Personally, I only questioned this as I’m looking for a new car and was shocked at the lack of fuel efficacy improvements of my current 2010 Toyota Prius which gets virtually the same fuel economy as the 2023 model, and it made me wonder what would the impact be if instead of switching to autonomous electrics, we simply improved how we are currently using today’s vehicles.
This future means that cars would be consuming far less gas, think UPS only making right turns or roundabouts that have already substantiated the savings, all without having to switch to electric vehicles. This concept also brings with it serious privacy debates and a political risk of being totalitarian, unless of course it’s a huge success and the system doesn’t penalize a person for spilling coffee and veering off course or said less politely if it could allow a socially acceptable level of failure to account for human stupidity and automatically adjust accordingly to balance human sentiment with performance. Yet, this failure risk, project duration, and a myriad of other factors are why it’s easier to stay in the periphery and the path of less resistance is to not question the narrative, progress must be good, and eclectic cars, unlike my 2010 Prius are cool! Disclaimer, I am not against electric cars, I have not found the study that shows what the carbon footprint reduction would be if all cars drove at optimal rates and I am not disagreeing that electric autonomous vehicles are relevant technology for the future, simply saying there may be parallel advancements that could also have a dramatic reduction to our carbon footprint while improving our quality of lives at a lower cost and quicker turnaround. Transitioning back to construction tech, the point that I want to highlight is the susceptibility of these shiny objects to becoming surrogates for the goals and that once a narrative is established it can be extremely difficult to switch directions.
Construction, like the automotive industry, is massive and accounts for ≈4% of the GDP in the US (Automotive is ≈3%) and creates our shelter, our work, and our entertainment. We’ve seen venture investments in ConTech growing from under $0.1B in 2011 to $5.4B in 2022, while still trailing the $10-13B invested in electric vehicles and components, it has become a formable industry, yet fragmented and lacking the generalizability to fully capture the excitement and investment like automotive. And while nothing exists in a vacuum and urban planning, construction, mobility, etc. are all interconnected, the ecological impact of construction and its product, buildings, is comparable with energy use in buildings accounting for 17% and transportation 16% of total CO2 emissions (noting these figures vary substantially to what’s included, how they are calculated, etc.) This oversimplification of the macro-picture is simply to illustrate that while these industries are comparable in size and impact, automotive receives a disproportionate amount of the global discussion from investment to regulations and that ConTech has potential for growth. A myriad of reasons can easily explain this from the fragmentation of the construction industry lacking lobbying groups to cars being a consumer product, or maybe it was just an evangelical entrepreneur who parlayed his success with Paypal to forge a new car company that capitalized on America’s love of cars with the perfect storyline. (Tesla does not define itself though as a car company) Whatever the reasons, the fact is that ConTech is here, it has its storyline(s), its rising unicorns, and my question of all is rambling takes me back to highlighted susceptibility above, which has modular; 3D printing; construction management software; BIM; Digital Twins (reality capture); and other technologies become the goal with the assumption that their growth will bring the promised productivity gains; automation / reduced labor demand; and provide a more sustainable construction?
Up to now, I’ve loosely established that ConTech can be both economically and ecologically with near parity to automotive, yet the funding is not as universal as is within the automotive space. Continuing with the analogy of the switch towards EVs, there is the question of infrastructure, i.e. building out a charger network. Construction is sitting on the opportunity to standardize its operations and capitalize on the arbitrage of its lack of productivity increases (almost every pitch deck has some form of the Mckinsey productivity chart from 2015). As the nation starts to cut deals to open charging networks and set standards on universal charger designs, the myriad of ConTech advancements have done little to standardize permitting, building codes, take-offs, bidding, and project management despite the availability of tools like BIM, LEED, and a myriad of project management software for years. This lack of transparency and mention of the bureaucratic costs associated with construction makes it challenging for new technology to be adopted and scaled into the built world. The bespoke nature of buildings means per-project set-up costs easily negate the potential savings and even if savings can be achieved they are benchmarked to unrealistic expectations like not accounting for change orders or unexpected permitting challenges. This lack of tangible productivity gains is a narrative though that gets little voice as speaking out goes against too many vested interests, from VCs that hype up their investments (primarily in software), to corporates that want to highlight that they are progressive, to the self-fulfilling world of ConTech with its conferences, to the start-ups themselves that have to sell the dream. “In theory, theory and reality are the same. But in reality, they are different.” This quote which lacks attribution dates back at least to the late 1800s and speaks to me of the importance of predictability, knowing the real world infinitely better so that theory merges with reality, is in some ways paraphrasing the promise of the day with A.I. (Artificial Intelligence). Predictability is the slow and steady metric, that will win the race to increased productivity in construction, yet as mentioned, slow, steady, and interdependent makes for a lousy pitch deck, especially over the past several years of optimism. As every industry currently works to understand how AI will impact them, potentially now is the inflection point when ConTech needs to come to terms with the excitement of this recent boom as construction was the last frontier (maybe better the most recent frontier) where LPs, GPs, and others flush with cash where excited to repurpose their software to this industry in need of salvation. To be candid, I shared this optimism bias, it was exciting times to be in ConTech, yet looking backward, I caution that this is a delicate balance, and while I may have been misguided, I was not mistaken. Every project needs funding, they need favorable sentiment, they need innovative technological breakthroughs, etc., but this balance is a delicate one, as bubbles can over-allocate resources that contribute little to fundamental change and this lost momentum can cause a downward spiral in construction tech. So as the industry now adopts the latest trend (not to infer that this trend is good, bad, or forever, simply that this is the latest fad), I’d like to highlight my opinion that most often technology is most beneficial in an aggregated use case and this is an opportunity to reshape the narratives of recent.
3D Construction printing (3DCP), is one of these golden technologies, it embodies the propensities highlighted above where we speak more of something being 3D printed than its comparison to conventional. Justifiably, 3DCP offers the ability to be the pacemaker to a construction site, opening up the potential for the entire ecosystem to rely on its standard of truth to compress build times, as the as-built matches a digital file, which means prefab elements require no site rework and deliveries scheduled without hiccups or insurance companies now have a timestamped digital file to use in the actuary tables. The technology noted is still very much in its infancy and early scaling stage, so cost and state of the technology aside, the allure of printing has completely replaced its intended purpose of increased productivity, decreased labor, and overall cost reduction. All noted and easily excusable if the robotic precision was measured by some standard of predictability, like OEE. It’s completely normal for start-ups to have a delta as the prototype’s features and benefits are still in development, simply noting that this delta is extremely well hidden with 3DCP. Nonetheless, this golden standard of OEE does not solve everything, as autonomous vehicles for example that have been increasing reliability to detect dangers receive press coverage of every accident while nearly 100 people die daily in the US from car crashes, which highlights this journey of tech adoption is not without a barrier of human sentiment, fear, and acceptance. One example that speaks to the success of prioritizing predictability fits both the bespoke nature of construction and the power of user adoption. In a talk with founder Uri Levine, he revealed that the algorithm was programmed to highlight the route with the highest probability of arriving on time and that the maps were user generated. The user-generated component means that if the majority of cars turned in a circle that it would infer a roundabout and this is how they generated the maps, while I’m not an expert in project management tools, the ones I’ve used have all been top-down management, not ground up like this (no pun intended). The second component was prioritizing predictability, this may be counter to most people wanting the quickest route, just like construction we want to improve speed, but this “guarantee” of sorts led users to believe in the quality of the product, thus bringing up user adoption, and eventually having robust options to also provide the quickest routes. I have not seen true empirical data yet on user adoption and documented benefits of a lot of the ConTech software that has been rolled out in the past years, yet confident to speculate that few have this compounding rate of increasing predictability. As this theme of unicorn stories of singular mutually exclusive technologies like 3DCP continues to grow, we also run the risk of being blindsided by parallel advancements with incumbent technologies. There are a myriad of digital fabrication options from automated CMU placing robots to autonomous drywalling and more, and without a proper baseline to connect these, I fear the construction industry will struggle to meet its innovation targets.
As the ConTech industry now begins to integrate and adopt A.I. as the solution, it’s important to remember that no industry exists in a vacuum and one paradigm-shifting innovation does not negate a myriad of others. Construction remains a diverse and complex equation to solve, as years of processes have trained contractors to pad their bids and keep their guard-ups. Litigation from project delays, and safety issues, to failing buildings, has led to an industry cautious to embrace change, and rightfully so. Yet, the industry has made progress in getting publicity for good and bad, it has brought in funding, investors, and technologies from silicon valley, academia, the military, and more, minting unicorns, creating new university departments and curriculums, and the large developers and builders have all created innovation teams. As the macro-economy remains elusive, this is a time of realignment and I ask the industry to be critical of the narrative that progress is inherently good, and that sometimes vested interest may not share the same time horizon to honestly create the infrastructure change that is needed and that the industry demands empirical measurements of efficacy.
So while the ConTech community continues to capitalize on the arbitrage between current building techniques and Industry 4.0’s promise, I am also asking myself what the future of buildings looks like. Not to undermine the needed and already achieved strides in building technologies, but like Uber did not change the product (i.e. the taxi) it changed the service, the construction world also needs to start asking how it is changing the service? What will it mean when a building will have a complete BOM (bill of materials), its digital twin, where the building can be valued independently of the land (i.e. separate real estate into a building and land product). Will this create new markets for interchangeable building materials or components, as up until now the model was that the depreciation of the building was to be offset by the appreciation of the property. (Analogous to leased cars valued off depreciation or the used car market) Or will new use case models emerge, like with cars as autonomous vehicles become the norm, and parking garages will be repurposed as cars will be shared. The point I’d like to end with is simply that the construction industry is only at the very beginning of its transition and we really have little idea what the future of sharing economy, IoT, smart cities, decentralized production, and the rise of A.I. will impact how the world interacts with its product, buildings, but the one element that appears clear to me is a need for a better-packaged product. Buildings need to be standardized and digitized not just from the composition perspective but also from the modification and construction perspective so that these new technologies can speak to and reallocate resources, like building materials and building equipment so that like my vision of the interconnected roadways we can work towards a common goal. This goal is a journey and I hope that one day we’ll have both autonomous electric vehicles and the interconnected infrastructure, just like I hope that one-day construction will deliver a 100% predictable product that is ecologically sustainable and economically available to provide the needed shelter and productive spaces that we live in.
Postscript on the title:
This article illustrates some of the narratives that we, the construction tech world, have created and how they themselves can become truths. So I ask myself, how different really is our intelligence that is judged by our ability to evaluate technology based on these creations, so different from that of A.I… one main one is that we screen for and are critical of the data sets used to train computer models while paying little attention to where our own biases may come from.
About the Author:
Matthew Carli is a global construction tech executive with an accomplished track record of founding and spearheading innovation efforts from starting an innovation team to developing holistic operational strategies. With experience in both the start-up and corporate world, working and living around the world, an MBA in international business, and a background in strategy and business development, Matthew brings a holistic human perspective toward shaping new ventures. Passionate about creating positive change he currently consults start-ups and investors. Based in Miami, FL, and open to new opportunities.
matthewdcarli@gmail.com | https://www.linkedin.com/in/matthewcarli/