For roofing industry, instead labeling of manufacturing environment, it will recognize spots with problems including preliminary classification and if the specialist needs to check spot in depth.
So, technology is already here. It’s time to at least try it and build the team of professional who will convert opportunity to real money.
Now, after we got flavour of sunny future, let’s look at the roof from 50 feet. Funny enough, now, we can do it literally, with drones.
Due to low cost, easy control and high definition pictures/video capturing, drones are on the way to become an important part of construction industry. Roof inspections/evaluations/proposals are going to be done using this powerful tool providing high speed, safety and thorough process documentation.
Today, usage of drones raises questions of authorities’ permissions and privacy. Lack of regulations can’t stop using of drones. For example, the police already use it for monitoring, and soon we will see drones fully equipped for crime prevention. From the other side, almost every industry desperately wants to use drones for delivery, inspections, monitoring. Business use cases cover every imaginable situation. In comparison, roofing industry has couple of limited, simple and harmless use case. So, if someone manages to get general permission, it almost automatically provides the path to get license to use drone for roofing inspections.
How to start. Drone business application requires significant R&D efforts from beginning. So, better to use existing hardware solutions to acquire initial expertise and prepare all legal aspects of usage. Internally, it would be ideal to create team working with one of innovative constriction partner to get real-life experience and data. From IT standpoint, it’s obvious that company should have dedicated Big Data solution to manage video and tons of high resolution pictures. As roofing business has its own set of tasks from roof inspection to final evaluation/proposal, there are should be number of highly specialized Machine Learning algorithms to extract value from massive amount of raw data. It includes computer vision systems, algorithms detecting anomalies and smart integration with company ERP, MIS, CRM systems. Although, it sounds huge, it mostly requires combining of technologies already presented on the market. And yes, you need bright people here.
Next Stage. After initial value generation processes are set, and drones become part of practice, next step could be pushing some smart algorithms closer to inspection point. I see it as smart filtering and getting results directly on site using image recognition/classification technics. Economically, it’ll help to eliminate moving useless data back and forth, and to save money on storage/processing.
Clouds. On-site inspection information is natural candidate to upload/process into cloud solutions. Even today, it’s affordable to put to Amazon Web Services one site inspection information (10GB) and keep it for couple months for just 24c/month (as of 2017/10/18). My personal advice here is to start with Clouds in the project right away. Building appropriate scalable IT infrastructure is way beyond roofing industry budgets and capabilities. In addition, today’s Clouds have advanced security and reliability; it’s critical factor if we want to provide excellent experience to hundreds of inspectors interacting mostly through mobile apps.
Outcome. With drones, software solutions and trained people in place, roofing companies have number of benefits which can be evaluated strategically and from cost saving perspectives.
Strategic benefit is access to raw information about local markets, products quality (including competitors’), installation issues. Having/providing fleet of drones with training and expertise is great marketing point to deepen relationship with contractors.
Usage of drones is the easiest way to accumulate enough data for machine learning. If competitors ignore that fact, they will be years behind of building appropriate algorithms and value generating services.
Cost savings come from faster, safer (less climbing) and thorough inspections and problems autodetection based on video/picture. In addition, using drones to estimate damage after storms can be done fast, risk free and using company personal globally when local contractors are busy with emergency calls. In general, cost savings and reduction of time of inspection could be so significant, that it will be profitable to provide it as online service to end users.
Risks. Investment into drone technology can be relatively small. To start, you just need couple drones and small team of engineers and developers. In addition, to develop first service offerings, it’s possible to use partners, some tech start-ups looking for applications of their ideas. It has some risks in Intellectual Property area, so contracts should be properly designed from beginning. But, real risk here is government regulations. They can make using drones illegal and it will make the investment almost useless (except part with Machine Learning, which is universal).
Roofing Industry. Drones as Opportunity
To check that everything I describe is already real today, here is video from Qualcomm Research https://www.youtube.com/watch?v=VwU9pPMqJh0 . You can see how drones applies machine learning classifier in real time.