Challenges and Opportunities for AI & Robotics startups Selling Into Construction

We are deep tech investors. That said, for even the most paradigm-changing technologies in the world, a thoughtful go-to-market is critical.

In a number of cases we’ve encountered, despite having groundbreaking technology, founders haven’t looked deeply enough at the existing market dynamics and haven’t identified a clear pain point they can solve in the short term or how to get to the defensibility they need in the long term. This is especially true for industries where there are multiple stakeholders who often have misaligned incentives, where the downsides of taking on new technological risks are rarely outweighed by the potential gains, and where the status quo culture is one that can be staunchly against change, particularly change brought by outsiders. In other words, it is hard to sell into the construction industry.

That said, there are certainly opportunities for technological advancements, especially in AI and robotics, to have an impact on the construction industry and we’ll outline some of these opportunities below.

As part of our “What to Build” series, we’ll also be interviewing a set of industry experts and getting their thoughts as to where the biggest, yet sometimes overlooked, opportunities lie (see below for our interview release dates). We’d also love to hear from you if you have thoughts about this space — feel free to drop us a line at hi@bv.ai!

Difficulties selling into construction

But before we get there, let’s go through some of the key reasons for why construction is a hard nut to crack, despite its trillion-dollar tantalizing figures.

1. The uniqueness of each project makes it difficult for standardization and optimization

In few industries do groups of stakeholders come together for a specific period, complete a project, and disperse, potentially never to work with each other again. Every project differs depending on where and how it is being completed. Different functional and labor constraints and requirements mean different outcomes. Each project also comes with its own set of expectations and process preferences from owners, developers, and other stakeholders. In other words, each project is unique, making it difficult for standardization and optimization.

In contrast, it is much easier to standardize and optimize processes in self-contained environments such as manufacturing. This is because the entire process is accomplished with the same set of machines and workers following repetitive procedures in an indoor environment that is easy to control.

2. Low margins require fast return on investment (ROI)

Despite the tremendous amount of money spent in construction (topping $1.2 trillion in 2017), net margins for contractors averaged only ~3.7% in 2018 (although this is higher for subcontractors). Few industry players have the luxury of investing into new technologies, especially if the return on investment isn’t proven or timely.

However, this can also be beneficial for startups. Some firms may decide to partner with innovative external teams if they can’t afford to create their own research & development departments (R&D).

3. Misaligned / tricky incentives among stakeholders

Incentives are not always aligned in construction and liability can sometimes feel like a hot potato that’s being passed from stakeholder to stakeholder. For example, if you’re selling a solution to the general contractor (GC): does this mean that the GC would also have to take on liability for any issues that arise? If the current liability is held by the GC and all the subcontractors, how accurate, precise, and robust would your solution have to be in order for the general contractor to willingly take over full liability? Concrete topic examples include layout and materials delivery, the responsibility for which is now spread out across the GC and all the subcontractors.

There can also be incentive misalignments within one stakeholder group for your product. For example, if you’re building a software solution to help architects ensure code compliance, you need to understand what the potential upside and downside scenarios would be. Architects are legally responsible for code compliance. In the worst case scenario, not following these compliance rules can lead to the demolition and rebuilding of non-compliant aspects at the owner’s expense and/or injury and death for construction workers; in both cases, lawsuits will likely be directed at the architect and could mean the loss of their license. Does your product’s upside sufficiently exceed the potential downside? What is at stake?

4. Lack of existing data hindering product or service build-out

Depending on the type of solution that you’re providing, you might need access to historical datasets that do not exist, are not sufficient, or require a lot of cleanup in order to be useful. There is also the issue of data being scattered across multiple systems that are not interoperable and that might also be difficult to integrate with. This problem is especially tricky for AI solutions that require a large enough training dataset just to get the product up and running. Since construction projects generally represent such a diverse array of conditions, it’s especially difficult to generate a representative dataset.

5. Distrust of outsiders, technology and change

While this doesn’t apply to everyone in the industry, depending on who you’re selling to the end users may not be eager to adopt new ways of working or new technologies. They may also distrust solutions coming from people who have not come from inside the construction industry. Construction is still very much a relationship-based industry and that can make selling into it harder for outsiders.

Stakeholder priorities within construction

All that aside, there are certainly opportunities within construction. To better understand where they lie, we have to look at each stakeholder individually versus seeing construction as a monolithic whole. Within the construction value chain, there are five major buckets that sometimes also overlap:

1. Owners/developers: The owner/developers are the ones who obtain the land on which a construction project will take place. Owners and developers can be private, government, or public entities, as well as others.
2. Consultants: Architects: are responsible for the look and design of the building. Engineers: create solutions to technical problems using mathematics and science. Surveyors: Determine property boundaries using exact measurements.
3. Contractors: General Contractor: provides all services, material, labor, and supplies for a project to be completed. Sub-contractors: Carry out the work for a company on a larger project.
4. Construction workers: Laborers who complete a variety of tasks on the job site.
5. Facilities managers: Ensure that the completed building is being properly managed and the tenant experience is up to expectations.

(Note: there are other categories including financial lenders, commercial real estate brokers, governments, etc. we have chosen not to focus on these. For this reason, we have excluded them.)

Various companies cut across the above segments. An example is Greystar, which is an owner and developer but sometimes does construction and building management; what Greystar does not do is construction work for other developers. In contrast, Simon Property Group owns, develops, and manages retail real estate properties but does not perform construction on any of its properties. Skanska performs development, construction, and building management, whereas Turner specializes exclusively in construction. Bechtel primarily handles construction but also plays a role in developing and financing specific projects.

Given the caveat that segments are not as clear cut as they may first appear, let’s go through what each stakeholder segment prioritizes and cares about:

Beyond merely understanding priorities though, it is essential to think through the cultural and behavioral implications or other constraints of any product or service offered to each stakeholder segment. Obstacles include everything from labor union regulations to the margins inherent in each segment’s business model.

For example, controlling construction costs may be possible only up to a certain point — for example, in regions such as the Bay Area, the scarcity of available land will always drive up the baseline prices being paid by owners / developers. Or there are only a small group of subcontractors who can handle the quality required in a certain region and they are not available to start on a project until much later than expected, causing a scheduling delay that would have been hard to predict.

Opportunities within construction

Opportunity areas for building each segment becomes much more evident when taking into account each segment’s top priorities. The most significant opportunities are ones that solve challenges for multiple stakeholders, naturally aligning incentives and ensuring that the solution to the problem is critical and reaches many people. The top three we’ve found based on our interviews and research are in these areas:

a) Avoiding delays in schedule and unnecessary rework

b) Worker safety

c) Augmenting worker capabilities

The opportunities vary within each of these priority areas, as multiple solutions can address each topic. Below we go through some of the main opportunities and dig a few levels deeper to grasp what’s beyond the low-hanging fruit. Please note, some of these ideas below address multiple areas.

1. Prefabrication and modularization

The continual rise of prefabrication and modularization goes in lockstep with all three priorities. It allows fewer workers to complete necessary tasks in a shop setting. This increases worker safety due to standardized processes. It also adds to the average worker’s capability because they no longer need to be an expert to understand how to assemble a prefabricated part. Performing well in this area will save both cost and time while preventing unnecessary re-work.

The movement towards a more standardized / in shop environment also provides opportunities for lean manufacturing techniques and technologies to be adopted. A lot of what has worked in manufacturing can now potentially also work for construction, which has seen low rates of productivity over the last decade.

2. Best practices knowledge augmentation

Given the labor shortage and with even more experienced supervisors retiring over the coming years, having a repository of construction knowledge that can provide the exact information that a worker needs at the right time can be powerful.

For example, uploading a BIM model and then having the system show the workers what are the common types of mistakes or edge cases that come up with this type of construction, or how to do this type of work in particular weather conditions, etc.

In general, any easy-to-use knowledge enhancing tool can be helpful for workers, especially those new to the industry.

3. Intelligent procurement of materials, and delivery management

Another opportunity area is in improving the supply chain and procurement process; this process includes tracking not just the materials from the suppliers to the warehouses and shops, but also into the field and potentially into the final structure for continued monitoring. Visibility into the supply chain should be as minute as possible to ensure that relevant workers know exactly where the materials are at all times and can even prepare for their arrival at a predetermined time slot on the construction site.

Over time, with enough data, the system should be able to make and even take procurement actions according to which materials are used most often for different types of constructions, and order them as needed while taking into account the suppliers’ lag time. This process becomes even more feasible with more standardized, modular prefabricated parts.

4. Coordination & communications tools

Technologies that can help with reducing coordination confusion among stakeholders will help keep a project on track. Currently, no such tool exists and each project relies on an amalgamation of different communications mediums such as phone, email, SMS, etc.

A real-time communications tool can also potentially act as a leading indicator that things are not going well: for example, if teams are not communicating enough or clearly with each other. Over time, algorithms can help compare communications and other indicators across projects, correlate that with the success of the projects, and hopefully come up with a monitoring tool that can track when a project is likely to be delayed.

5. Real-time environment data capture, understanding, and error reduction

Currently, construction planning is still primarily done with Gantt charts. Unfortunately, Gantt charts are not updated in real-time, and estimates of time for each task are often unrefined. These delays cause domino effects if the plan isn’t updated quickly enough.

It is often hard to analyze what is taking place on job-sites. Having an automated way to passively collect data on these job-sites can have a significant impact, especially with large projects. The hope is to use this data to better understand how the project is tracking versus the planned schedule, and to quickly take action if any delays occur. Our portfolio startup Openspace.ai is working on exactly this problem.

6. Document storage and management

Many parts of the construction value chain are bogged down by administrative tasks. This is especially true when it comes to document storage and management. For architects and contractors/subcontractors, the back and forth request for information (RFI) process can require a full-time person to coordinate responses effectively.

For engineers and project owners, the document submittals process can be equally burdensome as drawings, plans, and calculations have to be submitted in advance to the project owner for review and there can be hundreds of revisions of such submittals for one project alone. Currently, it’s a tedious process of making sure that if something is revised, everyone on the project receives a new version of the plan, which sometimes might only exist in manual paperwork form. While there are current solutions out there, the costs are still high relative to the value add for a lot of firms.

7. Data storage and management

Existing ERP systems often don’t provide enough of a friendly interface or data architecture to allow someone who doesn’t have a strong SQL background to easily pull relevant information. Designing a data storage and management system that can be easily used by anyone to create fundamental business analytics dashboards and to answer common data questions about on-going projects can augment existing capabilities and potentially ensure better project monitoring.

A separate but related issue is that one person overseeing a project may be leveraging a multitude of point solutions, downloading data from all them onto an Excel spreadsheet in order to make sense of what’s happening on a macro level. There is a need for a bird’s eye point of view dashboard that makes it easy for someone to see all of the data in one place.

What To Build in Construction

As with all solutions though, startups looking into the construction space should be aware that even within one stakeholder segment, every company is different. Some companies will be more willing to take risks by partnering with innovative early stage startups while others would prefer to wait until the technology is more proven out before engaging. For example, Swinerton would prefer to be one of the first one to three customers for a technology startup so that they can help to shape the final product / service. In contrast, Greystar typically only works with startup founders who have come from the industry and who they’ve had a pre-existing relationship with for many years. Choosing which doors to knock on when is key to not wasting your time or others’ time.

Speaking of which, as part of our What To Build series, we’ve interviewed a number of experts across the construction value chain. These experts were asked to give existing or potential founders ideas of what and how they should be building for the construction space, and how to engage with potential partners. We will launch a new interview every few days starting mid-November, in the below order:

1. Eric Law, Senior Director of Technology & Innovation @ Swinerton
2. Robert Kipp, General Superintendent Delta CM Team @ Laguardia Airport
3. Sharuk Khanna, Director of Innovation @ Disney Construction
4. Daniel Ramirez, Director of Innovation @ Skanska CDUS
5. Gaurav Joshi, Operations Manager Bay Area, Special Projects Division / Head @ TurnerLabs
6. Anya Boguslavsky, Project Manager @ Greystar Development
7. Atul Khanzode, Member of the Management Committee @ DPR Construction
8. Travis Voss, Leader of Innovative Technology @ Mechanical Incorporated
9. Eric Baldosser, Senior Project Architect @ Heller Manus Architects
10. Kitty Sullivan, Early Stage Investor @ JLL and Andrea Jang, Head of Growth, Americas@ JLL