Ziv Katzir

From Startup Nation to AI Nation: No Stone Left Unturned

Israeli companies are striving to discover what happens when AI is introduced into one of the most traditional of all sectors – the world of construction

Buildots was established in 2018 by three founders: CEO Roy Danon, Yakir Sudry, and Aviv Leibovici. The three completed their service in the defense system and decided that they want to utilize their talents and the abilities they acquired in the world of technology with the aim of changing the most traditional industry in the world. “We believed that it suffered from significant disparities and core problems, and we quickly arrived at the world of construction,” says Danon. 

The construction field presently comprises 10% of the world’s GDP – a huge industry with a tremendous economic influence. “The ability to build quickly and cheaply has a global impact on numerous areas,” Danon explains, “and we therefore decided to focus on it.” 

As people without prior knowledge of the field and without a specific problem they wanted to solve, the three decided to study the industry from the bottom up, asked the difficult questions, and quickly realized that a construction project is, in essence, similar to a production process in terms of working according to a designated plan, a schedule, sketches etc. However, once a process starts in practice, almost all the control and monitoring ability is managed according to a gut feeling or fragments of information, because there is no system supporting and accompanying the complex project. 

Danon describes a situation whereby in contrast to a factory, for example, a project engineer or site engineer lacks the ability to gain an in-depth understanding of what happens at the construction site and many decisions are made on the basis of unreliable, partial, irrelevant and even incorrect information. 

“Ultimately, the managers – however good and intelligent – make decisions on the basis of inferior information, making it very difficult to achieve good results. That’s what we set out to facilitate and resolve: to create a reality where construction projects are managed by people who understand exactly what’s happening within the process,” Danon claims. 

The assumption of the three was that the desired solution could not be human based because the quantity of data associated with a construction project is enormous. And that is exactly where they found the technological challenge they had been looking for. 

From 30% Goals Met to 90%

“It makes no difference to the construction industry whether a solution is implemented by people or by AI,” Danon says. “Ultimately, you need to receive a lot of information at any given moment. The difficulty is that humans cannot contend with such large quantities of information and that’s why we need to include algorithmics.”

The technological solution developed by Buildots creates a continuous connection between the building plans, a project’s schedule, and progress in practice. This solution is based on helmet cameras worn by the project manager, quality controller or someone else on the construction site, documenting everything that happens there. The information received from the cameras is gathered and analyzed by AI, machine learning and image processing. Artificial Intelligence enables to understand what happens in every room and the progress of construction in relation to planning, even though this is a noisy and non-sterile environment. 

“We can arrive at an accurate picture of what happened in the project in relation to the plan – down to the level of the individual power socket – and enable every project manager to understand the output of the team he is responsible for,” Danon explains. The results allow precise monitoring of on-site performance – to better plan the work and optimize the entire construction process. 

The final data is submitted to the client in a form they are accustomed to – reports that detail what has been completed and what has yet to be performed, down to the level of each individual building floor and professional. Buildots’ system allows not only to see what’s happening but also to manage the process practically and to integrate goals and weekly tasks into it. 

“This is a significant improvement because both the weekly plan and the results can be seen in an organized form in one place where planning is also undertaken. This makes it easier to adopt the various tools,” Danon explains, adding that “we already have clients who use the management system’s new capabilities, and its looks great.” 

The company has already completed several projects and, according to Danon, the companies reported that these were their most successful projects in terms of profit margins and performance. For example, in one case, one of Buildots’ larger clients used an industry metric that checks the number of tasks completed out of the project’s weekly plan. Before using the system, the metric stood at just 30% i.e., only a third of the planned tasks were indeed performed each week. In contrast, when the system was used, this figure rose to 90%. 

“One of the difficulties in managing a project is the ability to predict future events – its predictability,” Danon says. In other words, this reflects project managers’ ability to formulate a real plan that is actually implemented in reality. Buildots takes pride in improving predictability – something that could lead to a significant improvement in all construction metrics. 

Danon stresses that the system does not replace the humans who manage the project in practice: “we don’t make decisions about what to do but rather, integrate into existing processes, and upgrade the quality of data that managers receive by several levels. The difference between a mediocre manager and a good manager is the attention to detail and that’s what our system does superbly.”

After years during which the industry became accustomed to operating under conditions of uncertainty, Danon explains that the next stage is to adjust the work methods and the processes to achieve high efficiency. “It’s not enough to have better information,” he says. “People need to know how to use and utilize it. This is the fascinating and challenging process that we undergo with our clients – change management. 

The Advantages of High-Quality Construction

The company’s clients today are mainly large-scale construction contractors, primarily from Europe and the US, alongside limited activity in Israel. 

The beginning was not easy. As Danon explains: “This is an industry in which it was extremely difficult to raise capital – risk levels are high, sales processes are perceived as slow, and a field that is less known to the venture capital industry in Israel. The Innovation Authority took a risk and provided us significant support in our early stages. It would probably have been much more difficult for us to navigate the initial years without this added funding.” 

And what about the future? The goal is for every construction project to be managed in a modern, precise and controlled way: “We want to bring the revolution that has been implemented in the entire manufacturing sector to the construction industry. To change perceptions, tools, and to build efficiently. 

“There are numerous applications for data generated by our system and each of these is a world of its own. For example, we can work with insurance companies to lower the risk for construction projects and to improve the premium. We can reduce the interest that banks charge for financing because the risk for these projects is low. This is very significant. There are many things that can be done.”

And no less important: the ramifications of using the system go beyond economic saving. Once the process is more organized, there are fewer safety and quality problems, an assertation which, according to Danon, is corroborated by studies conducted worldwide. 

To Save Lives – But Also Time and Money

The Israeli company Ception was founded by two entrepreneurs: the CEO, Tal Israel and the Chief Technology Officer, Yossi Buda. The two have worked in the field of autonomous vehicles for nearly 15 years, some of them in the IAI (Israel Aerospace Industries) where Buda led the area of algorithmics while Israel led the field of software. “We established Ception when we discerned a growing need for an answer to the question as to how to operate vehicles in a more efficient, economical and smarter way,” Israel says. 

The new company’s field of activity is the operation of heavy mechanical equipment – heavy industrial vehicles, from steam rollers and forklifts, through loaders to bulldozers and giant trucks. The common denominator of all these vehicles is that they do not travel on the road but rather, are used for work at an industrial site i.e., on construction or infrastructure sites, mines, quarries, and factories. 

Operators of heavy mechanical equipment contend with three primary problems. The first of these is safety – these are very dangerous tools that move in a challenging and changing off-road environment. The operators’ numerous “blind spots” mean that both the operators and all the other people in the surrounding area must be protected. 

The second problem is that of the high costs associated with this field – operating these tools is expensive, and productivity today is limited and far from being maximized. This means that assimilating a new technology can save a lot of money. 

The third problem is pollution: “These tools are part of a polluting industry, but the lack of alternative means there is no other option but to use them”, Israel explains. “Nevertheless, using them more precisely will lower pollution and environmental damage.”

“The “Mobileye” of Heavy Mechanical Equipment  

Ception’s solution is to install a system that includes a GPS, sensors, cameras, a communications unit, and a screen for the operator in each one of the heavy mechanical equipment vehicles on the site. All the data processed in the end unit is transmitted in real time to two places – first, to the vehicle’s operator who receives automatic alerts about disturbances in the area, dangerous movements he is making, or suggestions for optimizing operation. 

The transmitted data is also uploaded to a server where Artificial Intelligence performs advanced processing. The result is insights at site level that are made available to the site manager, the safety officer etc. 

Artificial Intelligence is also used in an application for the vehicle’s operator via the screen installed in the vehicle’s cabin and in the management system. The use of computer vision and deep learning by cameras and algorithms enables a 3D mapping of the surroundings and allows to analyze the objects and the structure of the terrain in the work area such as dangerous proximity to a chasm, a safety rampart, or other hazards. 

These advanced capabilities enable to provide extremely precise alerts and to identify only genuinely dangerous situations: “An operator who works long hours and receives countless false alarms, will ignore them,” Israel stresses.

Artificial Intelligence also helps to analyze the data gathered from all the tools at the worksite and stored in the cloud and provide analysis on the general site level and not just that of the specific tool. The safety advantages are clear, but the data also allows to map the state of work at the site and to indicate discrepancies between planning and actual performance, inventory, analysis of the state of roads and improving maintenance. “There is an entire range of things that we can do once we know how to gather and analyze the relevant data at the site, which is the real AI,” says Israel. 

“We have a catalogue of applications, and we add new ones for our clients according to needs that we identify on the ground and from client feedback,” says Israel. “The real treasure for the company is the data gathered in the field. We work with very large clients right from the beginning. In other words, we don’t develop a technology based on data from the internet but rather, using information from clients worldwide. For example, Shafir Engineering – that is active in all engineering fields of quarries, factories, paving roads, and infrastructures – is not just a client but also our partner, one that invests in the company. 

In relation to competitors, Ception presents its core technology capabilities that excel in their accuracy, identification capability, and all levels of performance. “Our technology is very strong, and the AI is so precise that it allows to develop applications that don’t exist today,” says Israel. “The business model of a single hardware kit that expands with a catalogue of applications is very innovative – especially when we provide a client with access to a technological product in a very traditional field.” 

For the Transparent Workers

The company’s first workers were hired during 2019. In the time since the technology was developed, pilots have been conducted in Israel and overseas and today the company is in the sales stage. 

“The Innovation Authority has played a significant role in where we are today. It has been supporting and contributing to our R&D effort for several years already. The bottom line is that even in tough times such as the COVID pandemic, it helped companies realize their dream and vision, to hire employees, and to expand the industry in Israel. This is what enables us to remain a genuine high-tech nation, something was made evident in our own private case,” Israel emphasizes. 

The company’s main target audience is industrial companies – large construction or infrastructures companies, mining companies, and materials production companies – to which the system will be sold directly. Another goal to which the company aspires in the long term is collaborations with manufacturers of heavy mechanical equipment. 

“Our vision,” declares Israel, “is to achieve a revolution in this field of operating heavy machinery i.e., to save lives and also money for our clients. Unfortunately, workers in this field are transparent, and it is important to remember that they work in a very difficult environment and are exposed to high risks.” 

There will also be positive ramifications for the end consumers building a home or driving on a road. “It takes 2-3 years to build a road and it is now possible to shorten this by several months, to reduce costs, and to limit environmental damage – this is all very significant.”

This blog is part of a series of five articles in which experts from the Israel Innovation Authority explore of how Israeli Artificial Intelligence is spearheading global development in the field. For more articles:

About the Author
Ziv Katzir is the Head of the National Program for AI Infrastructure at Israel Innovation Authority, an independent public entity that operates for the benefit of the Israeli innovation ecosystem and Israeli economy as a whole.
Related Topics
Related Posts