Real estate, usually a laggard in tech adoption, is finding its groove
For 21 years till 2011, Vikas Gosain worked with the Indian army. He was a Gorkha Regiment infantry officer. He has worked with civilian organisations since, and currently manages security operations for Bharti Realty, a real estate developer. On a cold December morning, he takes this writer around Worldmark 2, a commercial and retail development near Delhi's international airport. This is considered to be a 'sensitive' area, when it comes to security.
In one room, hidden from the hustle and bustle of offices and restaurants, is a series of 24 screens. This is Worldmark 2's security operations centre. Feed from 314 cameras in the building are beamed to this centre live. Every person and car entering its perimeter is scanned and the images are processed here in real time. "We cannot have a blind spot," Gosain says, as he proceeds to explain the technology behind real estate in high security areas. Every vehicle entering Worldmark has to pass under a vehicle scanning system whose high-resolution cameras scan the underbelly of the vehicles for any abnormality in seconds.
Facial recognition technology is used to identify miscreants entering the premises. A database of criminals and suspects, received from Delhi Police, is fed into a software system. Its algorithm matches every face spotted by the camera against the data stored. A match of more than 80 per cent is shown as a pop-up on the screen in the control room. New facial recognition technologies use deep learning, a branch of machine learning in which the algorithm learns from the data it receives every minute on what to look for.
Gosain is now considering deploying an Artificial Intelligence (AI)-assisted video monitoring network. Again, a system that learns daily to automatically pick up abnormal behaviour in the crowd. For example, if these cameras spot a bag that is unattended for 30 seconds, they can trigger an alarm.
Securing real estate is just one of the areas where technology is being used. Real estate is usually a laggard in technology adoption. Most commercial developments and hotels, for instance, don't use automated under-vehicle scanners; they make do with manually held mirror-based scanners. A smart algorithm could check for abnormalities and more accurately match that with a database of the right specifications in a vehicle.
The bigger developers are, however, adopting solutions that leverage AI to the Internet of Things (IoT), in a bid to stay ahead of the game. The Indian construction sector is expected to total Rs 50 lakh crore by 2022, a CAGR of 16 per cent since 2018. Globally, construction costs are moving up and developers are also struggling with increasing wage bills, a KPMG report points out. "Real estate developers are heavily investing in proven construction technologies bearing potential to solve time, cost and skilling issues," the report states.
The Indian real estate sector is also characterised by low trust, particularly when it comes to the residential business. So, developers see technology as a means to deliver homes faster, and in a more transparent way. Gurjot Bhatia, Managing Director of Project Management at advisory firm CBRE, points to the different areas in a real estate development's lifecycle where technology has started playing a role. It starts at the drawing board with conceptualisation. Then comes design, budgeting, procurement of construction material and monitoring construction. "There has been a focus on health and safety as well. The last part is managing the quality of work and handing over the project," he says.
Vikshut Mundkur, Co-founder and CEO of HUVIAiR Technologies, shows a video in which two men, wearing helmets, fire a white drone, propelled by four rotors. The drone lifts up rapidly before it is manoeuvred over a real estate project that is under construction. It sends back a bird's eye view of the land, full of construction material and partially constructed buildings.
"The developer, the project management consultant or even the contractor has the same flow of data. They are able to take decisions a lot quicker, a lot better," says Vikshut Mundkur, Co-founder and CEO, HUVIAiR Technologies, a drone services company that uses visual data analytics to track construction progress. -- Photograph by Sudhir Damerla
HUVIAiR is a drone services start-up in Bengaluru, with Boeing as one of its investors. After the drone captures the video of a construction site, the firm uses visual data analytics to gauge the progress of construction. In technical terms, the company uses a process called photogrammetry or the use of photography in surveying and mapping to measure. "We process the images and generate 2D and 3D models of the site. Then we apply computer vision, visual data analytics, and deep learning in combination to generate intelligence," Mundkur says. "The progress of construction is captured in a dashboard. Our algorithm compares the scheduled data with the visual data captured," he adds.
This technology helps avoid human errors in reporting. The common dashboard brings all the different stakeholders in a development on the same page. "The developer, the project management consultant or even the contractor has the same flow of data. They are able to take decisions a lot quicker, a lot better," Mundkur says.
Analytics is also helping developers with more efficient budgeting. A budgeting software from real estate enterprise resource planning (ERP) company BuildSupply, for instance, allows a developer to make a budget using a lot of historical data.
All these developments are, however, work in progress. "The current level of technology takes out a lot of the paper work, and having multiple people doing duplicate work. As we build up the data, we will be able to apply machine learning. I can start doing predictive analytics around what is going to be my cost structure in various geographies," says Sameer Nayar, Founder and CEO of BuildSupply.
A more established use case for machine learning is conversational AI or the use of chatbots and speech-based assistants to automate communication. Such systems use Natural Language Processing technologies. Here, machine learning helps algorithms interpret human language. Real estate developers are beginning to flirt with this idea. Godrej Properties, for example, is experimenting with a chatbot that can respond to user questions in text as well as voice. The bot can pull out and display graphs as well as data when needed.
"There is a paradigm shift in real estate security. From perimeter security to under belly scanners, everything is technology driven" Sushil Kumar Sayal, MD and CEO, Bharti Realty.
Vineet Bhardwaj, IT Head at Godrej Properties, says that the system's decision engine is connected to the business databases. "Conversational AI system is trained to learn business intent, synonyms, and context," he says. Initially, the chatbot is trained with a predefined set of questions and answers but learns the context better as it interacts more with users. "The system will either provide an accurate answer or prompt the user to provide more input or corrections to the input," Bhardwaj says.
Like in any other industry, chatbots do away or reduce the need for more humans in call centres, thereby slashing costs.
BuildSupply recently started a managed marketplace for real estate supplies. The company aggregates demand, say, for steel, and then decides who to buy from at the most competitive price. A real estate developer or a contractor depends on BuildSupply for both quality and price. To monitor quality of the supply, the company is banking on IoT; it has installed sensors in some supplier factories.
Making ready-mix concrete, for instance, involves several elements such as gravel, sand, water and cement, among others. Every batch of concrete produced has a design mix - specified amounts of cement and sand - which is determined by the structural engineer. Sensors in the plant monitor this mix for every batch that is made. "A lot of cheating happens at this stage. But we also monitor moisture content," says Nayar. Information captured from the sensors is uploaded to a Cloud platform from where it can be accessed through an app the company has developed.
Next, using GPS tracking, the company monitors the time it takes for a truck from the factory to deliver the concrete at the construction site. "Concrete can only be delivered between 30 and 40 minutes. The minute you pour the concrete, it starts setting. That distance and time has to be captured. This information is available on a real time basis to the buyer. So we have sensors in the concrete making plant as well as in the truck," Nayar says.
Although this use case appears relatively new for real estate, GPS tracking technology is fairly common in the logistics world today where it also solves the on-route visibility problem. Using GPS, IoT, and Cloud technologies, it is now possible to know the real time location of the trucks and track their movement. Many technology companies have now developed systems with a predictive engine. Based on the source, destination, load, type of truck, and number of drivers in a truck, the system can predict the ETA.
At Real Estate Major Puravankara's Factory in Bengaluru's Kanakapura Road, With the Use of Pre-cast Technology, Construction Time Can Be Shrunk 50% Versus Conventional Ways. -- Photograph by Sudhir Damerla
IoT, meanwhile, also has a use case in safety and security in real estate. Not just in commercial properties but also in residential areas. Property developer Mahindra Lifespaces says that it doesn't want to just bank on scanners and a network of cameras for home security. The company says select homes it develops can come pre-fitted with a range of sensors that cover gas leaks to glass breaks.
"We are exploring if we can offer these in our coming launches. Gas leak sensor will trigger an alarm; it will send an SMS; it can trigger a phone call at a nearby call centre; it can trigger a phone call to the home security. If I am part of the building security, I will be able to see on a building map in which apartment the alarm is coming up," says Abhinav Aggarwal, Head of IT at Mahindra Lifespaces. Data from the sensors goes to a Cloud platform and then to a network operations centre where it is processed and acted on.
"Controlling lights and fans (through IoT) is passe - they lose charm over time. We want to focus on safety and security," Aggarwal stresses.
The website of FlipSpaces promises "tailor-made workspace in eight weeks" and the fact that one can "see it before you build it". The design and contracting firm uses virtual reality (VR), a computer technology that creates a simulated environment. In other words, designers at FlipSpaces can help customers with virtual walkthroughs of office interiors they intend to design. Such walkthroughs are immersive, and the experience quite gamified. So much so, that it is a near-real experience of how the space would look after it is furnished.
"You can visualise the entire office or the retail space. You can sense how the office would look like," says Kunal Sharma, Co-founder and CEO of FlipSpaces.
Vizwalk, the company's VR-application for interior design and visualisation, is based on Unreal Engine, a suite of tools that game developers use to design games and simulations. Files with technical drawings with detailed floor as well as furniture plans are shared with the system. The app processes the files and generates walkthroughs. "In Vizwalk, you can navigate through buttons you would use in a gaming software. You can move around the office. While you move around, you can change the finishes of the space in real time like the wall or the flooring," Sharma says.
Apart from AI and IoT, virtual reality is one technology that developers and project management consultants seem keen on. Immersive experiences during the design phase help in faster decision making and in some cases, is easier for occupiers of commercial offices to understand and plan.
Advisory firm CBRE has come up with Floored Plans, an interactive real estate marketing and leasing tool. It can help one visualise and choose the commercial space to lease. There could be two 200,000 sq. ft building spaces in Gurgaon with the same rentals, for instance. How would a company decide which one to go for?
"Usually, an architect gives you a plan and tells you which building is more efficient in terms of capacity utilisation. Using Floored, one doesn't have to go to the architect," says Bhatia of CBRE. A user can feed in the requirements and the tool would automatically generate layout plans in seconds. One can also play around with the layout and customise the office theme. "These happen at the click of a button. The initial decision to choose a building and how the office should look can be done faster. Traditionally, such a decision can take three-four days. Here, you do it in hours," Bhatia adds.
There are other applications of VR in real estate as well. VR models can enable contractors or project managers plan various stages of construction; simulated walkthroughs, mostly in experience centres created by developers, help homebuyers visualise their apartment and is often used as a pre-sales gimmick.
While digital technologies are currently riding an adoption wave, there is an equally key construction technology that is expected to gain traction in the coming year - pre-cast - where much of the building can be made in factories by machines and then transported and assembled on the site. As a technology, this is decades old. Singapore, for example, has been building entire residential apartments on pre-cast for many years, even the entire toilet, complete with tiling, commode and mirrors. Indian developers have been shy of using such technology.
Sobha Ltd, one of the earliest adopters of pre-cast in India, points out that there are challenges in design, and in the architectural and structural elements due to construction joints and connections. Further, pre-cast requires high upfront capital investment and comes with limitations of erection, modifications and transportation.
Heavy investments, however, are better than facing a trust deficit from customers because of delays and penalties. With regulations such as the Real Estate (Regulation and Development) Act, all developers and promoters are realising that it is critical to deliver on time without compromising quality.
"Pre-cast relies a lot on software and technology. All the plans are loaded to computer systems and then at the factory, slabs get cast by robots. We are able to standardise quality and control time," says Ashish Puravankara, Managing Director at Bengaluru-based Puravankara Projects. The construction time, he adds, can be halved versus the labour-intensive traditional way. "If I increase my factory capacity, I can do it in one-third the time (versus conventional means). But then you have to find the right feasibility between cost and benefit," Puravankara says.
A pre-cast factory has pallets, which are like large tables. One of Puravankara's pre-cast factories constructing for a project called Park Square in Bengaluru's Kanakapura Road has 26 pallets, 12 metre by 4 metre in size. Steel shutters are fixed on the pallet to make a mould, say, for a wall. The plumbing lines and electrical junctions are fitted next. Finally, concrete is poured into the mould. Once the concrete settles down, the wall is removed from the mould. Cranes are used to take the wall to the construction site to be assembled.
In advanced factories, robots also draw the dimensions of the shutters while machines pour the concrete and can even remove it from the mould. The entire operation can be automated, and thereby, reduce cost if planned and executed well. According to executives at Sobha, conservatively, precast technology can lower the construction cost by 10-15 per cent. There is minimal dependence on labour and because of the factory setup, there is minimum wastage of materials on site as well.
Technology is changing real estate development and management like never before. As it starts becoming the USP for projects, developers will need to adopt it in all its forms. Those dragging their feet may find the going tough.