What To Build: Sharuk Khanna (Director of Innovation, Disney Construction)

Conversation with Sharuk Khanna, Director of Innovation at Disney Construction. We chat about ten detailed startup ideas for the construction space, what it takes to go from pilot to production, and the wide range of personalities in this industry and why one needs to take that into account.

1. Tell us about yourself, where you currently work, and your path on getting there.

I’m the Director of Innovation at Disney Construction. Disney Construction, no relation with the Walt Disney company, is a civil heavy general contractor focusing on primarily public works projects with some private projects as well. We generally work on bridges, highways, waste water, structures, etc.

Prior to working in operations, I worked as a project engineer on wind and solar farms. My experiences have exposed me to the numerous redundancies in our space and I see a lot of opportunities for optimizations and technology to change the way that we work. I’m focused on assisting our industry at a higher level.

2. Tell us about your role and what your mandate is and how this specifically relates to working with startups?

My mandate is to identify our pain points and priorities, and then find and evaluate products in the market that can help us. Since we don’t have the funding to build a research and development department in-house, we need to look for more off the shelf solutions and partner with innovative startups.

As part of my due diligence, I often loop in the relevant internal stakeholders at Disney Construction to help me assess the potential solutions via product demos. We then analyze the pros and cons of all the proposed solutions. If we find one that we really believe in, we’ll talk with our finance team to find the necessary budget. During implementation, we’ll get the relevant internal stakeholders onboard with training and rollout; these internal champions then help train up the rest of their teams.

3. What are some of the interesting types of projects that you’re currently doing with startups?

We’re currently working with Stanford in partnership with their engineering program to bring students on site to one of our projects and explain the relationships between schedule, planning, and building. They take an element of the job and recreate it using ALICE a 5D modeling tool and produce optimized work plans that could increase our production rates or provide relief in our schedule.

On a grander scale, I’m building a construction innovation counsel board with 15 other folks from all sectors of the construction industry in order to knowledge share about solutions that have worked across company lines.. We’d like to also consolidate a list of industry pain points for the startup community so that they can better understand what we need built.

4. What number of these projects move into production? By what criteria? One of the challenges we see startups facing is how to move a customer from pilot to production.

Thus far, we’ve done four pilots. In reality, we try to stay away from the traditional “pilot” model in that we want to bring in technology that we can easily ingrain into our culture and work processes; we don’t want to test a huge number of solutions and confuse our employees, who have to try different tools from project to project. We’d prefer to implement a few reliable solutions across all of our projects so that our workers only need to learn how to use one set of tools.

To go from pilot to production with us, one of the main things we look for is a strong training and support program. Making an upfront large investment in 24/7 customer and training support is critical because our industry has night and weekend shifts along with high worker churn — if something goes wrong, we need to be able to call someone immediately. You can’t have one account executive at a startup who goes home at 6pm and who won’t respond to emails until hours later. Having a robust support system will go a long way when it comes to evaluating products to bring into our operations..

If you’re going to be working with us on a construction project then you’re just as much a part of our project as any other subcontractor or critical job member and we’ll hold you to those standards.

The longevity of your product is also important for us. If it doesn’t look like your solution will be a standalone product for long — for example, your product looks like the type of feature that Autodesk will acquire to add it as a feature to their existing product suite then we might be more skeptical because we don’t want to depend on a product that will likely go away soon.

Lastly, working with founders who really understand our industry and who can quickly iterate are huge bonuses.

5. What are the major challenges in your industry these days, and specifically ones that you think can be addressed by the right type of AI and or robotics application? Can you give some detailed examples?

I have lots of ideas for challenge areas that should be addressed in construction! Some below:

a) Passive data capture & AI analysis: We need startups such as OpenSpace.ai to help us do better passive data capture. Eventually, AI should be able to tell us how many man-hours producing a certain structure would take, and couple that directly with costs in the backend accounting system. Another method is to create an automatic time card with a very simplistic UI / UX such that workers can quickly flip through a list of ten or so cost codes relevant to that workday and help them to track time; this would allow us to get rid of our existing time cards, which require the foreman to step in (versus workers doing it directly).

b) Robotic survey/layout: having a roving robot with a laser that can be remotely controlled and mark layout points within a quarter or half inch of accuracy would replace the one-man two-instrument teams that are currently needed for this type of work. This would reduce multiple days in the schedule and hopefully produce better accuracy. The difference between layout and surveying is that in layout, one needs to put points on whatever surface so that people can build projects off of such guides whereas in surveying, a specific geographic area’s topography needs to be understood.

c) Pre-Site survey with drones/ground bots: I imagine that we’ll likely get to this solution pretty soon. Drones work unless there’s trees, so that’s why you need to have ground bots. In the ideal scenario, you’ll only need one person to deploy both the drones and the ground bots for each site and then combined all of the photos together to get an accurate topographical model.

d) Streamlined bulk material ordering & delivery: we need a smart enough system where I can input what I’m building and its location and there’s enough standardization of the structure that based on the type of home and location, it can recommend me the most used materials. If we can get this down to simply needing to upload the BIM model for the structure and then the system can tell us exactly what materials we need to build it then we have Amazon’d construction.

e) Alternative ERP solution customized for construction industry: The ERP system is the hub of where data lies right now. The current players are Viewpoint Vista, Sage 300, etc. with some people using QuickBooks. Most solutions are not good enough to create a business intelligence tool that a construction person can use; they tend to do a poor job of document management and data storage, making it hard to pull meaningful data. We need a tool that makes it easy to understand how the data is stored and how to pull the data. Then can have useful dashboards. ERP systems are so critical because it’s a tool that encompasses all aspects of construction, from change orders to documents.

f) Business intelligence for construction: Related to the above idea, we need simple to use data tools that would allow someone with no knowledge of SQL to be able to do basic data analysis (ex. cost analysis, outliers analysis, etc.) and do basic data auditing or adjustments.

g) Dynamic detour planning / traffic control with location data: The current method of figuring out detours involve people counting cars over a period of time in order to gauge volume. There should be a better way to adjust real-time for traffic flow. Having this information is good for drivers and also for construction crew who need to get to the worksite on time, especially for night shifts, which are more costly.

h) Virtual Superintendent: since we have a labor shortage, we need to keep labor on site and not running off site to do menial tasks. We don’t have a consolidated knowledge center that teaches construction workers how a bridge or a building gets built, what to do when XYZ common or edge issues come up, how to install ABC in the winter-time, in dry areas, on slopes, etc. Having a repository of best practice knowledge would be a time-saver. One way is to collect all of this tribal knowledge from superintendents in pre-recorded sessions or to have retired superintendents consult as part of a community.

i) Work planning tool: this currently exists as a simple process in construction where knowledge about how to build a piece of a project is collected in such a way that anyone who receives the information should be able to read it and build it as you have. However, right now it’s a hodgepodge of internal conversations, spreadsheets and plans, typically collected in a 30-page binder that no one reads. There’s no tool that allows you to aggregate all the information and plans in a standardized manner.

j) Preconstruction automated submittal log: Automated submittals log — Whenever you start a project you get special provisions (i.e., standard specs of how to build in your state, then also special provisions for each project). There’s a process called submittals where you need to submit drawings, plans, and calculations 60 or 90 days in advance to the project owner (CALTRANS for us) for review, especially if they are engineering drawings. You might get revisions (i.e., requirements) on your submittals; there can be hundreds of submittals for one project alone. Currently, engineers need to use OCR to parse through hundreds of pages of documents to figure out the what and where of these revisions. Basically, if something is revised, everyone on the project needs to have a new version of the plan, which sometimes might only exist in manual paperwork form. Not only do we need to get people onto digital plans, we need to help save a lot of preconstruction man-hours and on-site construction time from workers (sometimes even one full-time person) managing the middles process. While there are solutions in the market for this problem being provided by PlanGrid and other providers, the costs are very high, in the tens of thousands of dollars per year.

6. What type of startup would you be most excited to see?

I’d love to see a larger version of Built Robotics that produces not just autonomous bulldozers and excavators but also an entire self-driving fleet of construction equipment. You can probably use such technology to do ~70% of the work without human intervention, including excavation, material delivery, ordering & pouring concrete, etc.

Ideally, one just has to upload a BIM model to pour 500 cubic yards of concrete in a 10 feet by 10 feet hole against XYZ coordinates. The software then dispatches a crew of autonomous equipment to drive to the coordinates, to dig the hole, to order and then use the concrete, and to pour it in with minimal human interference.

I think we’re not yet at the stage for the above idea but in the near-term, another more accessible idea for a startup I’d be excited to see today would be a streamlined bulk delivery startup. This would allow us to get all the materials we need at once, reducing the need for multiple trips and staging areas.

7. What should startups know about your industry before going in? What nuances or details about the industry are not so apparent from someone looking in?

You need to be receptive to a very, very wide ranging set of personalities in our industry. On the one hand, there are people who don’t even know how to use a computer whereas someone else is a true techie and your solution needs to be able to appeal to both of them.

Construction is a very time limited industry — we make money when we build during daylight hours (with a couple of night shifts). We don’t have the time to spend 10 hours a week for 3 weeks learning how to use your product.

Also, keep in mind that the high number of cross-functional teams in construction is greater than in other industries. There are so many segmented stakeholders in our industry that it becomes hard to manage all of their demands. But everyone is working towards the common project goal and the smallest role is as important as the largest role. It’s important to not take that for granted — your product needs to be good for the overall project.

Lastly, understanding the language of the industry is important. If you understand the process of what it takes to build a project, using that knowledge to inform how you build your product will ensure that you are providing a solution for a pain point that actually exists. Many people will be skeptical of your product if you are looking to us to show you how we would use it.

8. Lastly, any recommended resources / reading (ex. Industry conferences, publications, experts to follow, etc.) for startups looking to build in your space?

Go to industry meetings like the quarterly events held by AGC and UCON, as well as user group meetings held by Procore and HCSS. There you will see many people talking about the problems we are facing and how our problem solving differs from others. It is a good place to understand all the stakeholders within a project and the industry.