|Page (1) of 1 - 07/31/17||email article||print page|
The rapid development of artificial intelligence (AI) technology means it is quickly becoming suitable for a wide range of commercial applications, and civil engineering is one of the sectors where it could soon make a big impact.
Currently one of the least digitized industries, there is huge scope for AI to transform processes across the entire project life cycle. Such a significant transition may not be entirely smooth, but as the benefits become clear it is surely inevitable - so lets take a look at what the future holds.
In order to plan new civil engineering projects, we first have to fully understand the nature of an urban environment and how it is used; a process that normally involves 'spatial network analysis'. And as those environments become ever more convoluted, the software tools that perform that analysis are beginning to incorporate AI to handle the complexity; tasks which would have once taken humans days can now be completed in a matter of seconds. There is even the possibility that in the future citizen data could be fed into these planning processes, and deep learning algorithms used to uncover what new development would be most beneficial to the local population.
Moving on-site, and AI may now be used to analyze terrain ahead of construction projects. Geotechnical engineering, the study of soil and rock which will support civil engineering systems, is essential for laying effective foundations but extremely challenging due to the variability and non-linear behavior of the materials involved. However, it has recently been shown that machine learning can model this complex behavior with superior predictive ability, providing a powerful new tool for engineers.
The advent of drone technology will also make surveying more efficient, because the sensor data collected by UAVs (Unmanned Aerial Vehicles) can be used to create topographical maps of a site. An estimated $6.4 billion is currently spent annually developing drone technology around the world, and much of that is focused on developing it for commercial applications. And as the incorporation of artificial intelligence technology enables drones to work more autonomously, they will provide a faster, cheaper alternative to human surveyors.
Architects are already employing artificial intelligence in their design tools too. Interestingly, one current trend is to use game engines such as Unity 3D, but software specifically designed for architectural purposes has now been developed as well. In fact, some even use AI to automatically organize the layout of spaces within a building depending on the requirements and data supplied, speeding up the design process and leaving the architect free to make more aesthetic adjustments.
As in the case of geotechnical engineering, because of the complex nature of environmental systems it is clear that AI will play a key role in ensuring structures are well-suited to their surroundings. Not only can architects now conduct environmental analyses from their computers, using simulations to ensure that energy efficiency is maximized, but AI algorithms also have a big future in engineering design. Artificial neural networks can be used to select the most appropriate structural systems depending on wind and seismic loads, while genetic algorithms have been employed to optimize the life cycle cost of buildings in hot climates.
It is of course important to consider how artificial intelligence applications in a wider sense will affect the design approach as well. The architects of the future will need to factor in the presence of a multitude of IoT sensors, and depending on the situation, perhaps even areas where users can interact with a structure as an intelligent entity. Machine learning techniques are being used to both identify and synthesize new materials, which could open up whole new possibilities for design as well.
In Japan, an ageing workforce and labor shortages have lead to the country becoming world leaders in construction site robots. Unmanned dump trucks, bulldozers and vibrating rollers are already being used to improve productivity, and while they are currently operated by a human via a tablet, it is not too much of a stretch to see them working on their own as AI technology advances. Construction sites are particularly challenging due to the variability of the environment and the tasks required, but with the world's first commercial robot bricklayer even set to be released this year, the trend is definitely towards automation.
Machine vision powered by machine learning will not only help reduce the number of defects in the manufacture of building components, but will also help engineers to check the integrity of structures throughout the construction process. Drones could be used to scan sites frequently and help direct resource deployment, measuring progress through 3D modeling and monitoring stockpiles of materials through volumetric measurements. The key thing here will be to not just throw more data at project managers, but to find a way to analyze that data automatically and produce actionable results, and AI analytics may hold the key to that.
Planning of a project itself obviously begins much earlier in the process, but construction is one of the areas where AI-informed project management methods are likely to have a big impact. The use of genetic algorithms to optimize both the scheduling of construction site activities and the layout of the site are topics of recent research, while a range of techniques from artificial neural networks to fuzzy logic may be applied in cost estimation, risk management and performance management. Better project management can only lead to construction projects running more smoothly in the future.
In a similar way to which AI can be used to model terrain behavior in the planning phase, it can also be used to monitor the state of foundations after a structure has been built. Bridge scour for example, the removal of soil and rock from around bridge structures due to flowing water, is the leading cause of bridge failure around the world, but inspection of those areas is time-consuming and costly. Over 20,000 bridges in the U.S. are currently rated as 'scour critical', and by using AI to analyze the data from sensors and model further erosion, we can find out which are most at risk and prioritize repairs.
AI analytics will also be used to monitor the health of the structures themselves. For example Kone, a leading escalator and elevator company, has partnered with the IBM Watson IoT platform and will use its cognitive abilities to inform elevator maintenance. By analyzing the data from IoT sensors, Watson can predict the condition of equipment and any potential problems, and even suggest solutions - as this concept is proved, it will likely become a feature of larger-scale projects as well. Finally, autonomous drones will again play an increasing role; scientists at Carnegie Mellon University are currently developing 'flying bridge inspectors', which can scan for structural problems without putting human inspectors in dangerous situations.
So should civil engineers be worried about 'robots taking their jobs', as is the case in many other professions? In short, in the short to medium term, no. Although many tasks currently performed by humans will be automated, improving project efficiency, human engineers will remain, down particularly to the amount of creativity required in the role. Is it possible that human civil engineers could eventually be replaced? Yes. But there are many, many barriers for AI technology to overcome before we get there. Of course, we have already seen how in some cases robots can do the job of a construction worker, and some of those positions are certainly closer to being automated, but there will still be a place for the more artisan crafts and operation/supervision of on-site robots.
What is true is that the necessary skill set of a civil engineer will surely change. Working alongside artificial intelligence, they will be required to operate and interpret the output of those systems, so as with most industries, training will be a lot more tech-focused than before. Project managers, many of whose day to day activities are actually most under threat from automation, will have to refocus on the skills that AI doesn't have (at least for the foreseeable future), such as stakeholder management, interpersonal skills and emotional intelligence. Going forward, the most successful in the civil engineering sector will be those who recognize and harness the power of AI to allow them to focus on other areas, rather than trying to beat it at its own game.
Tim Scargill is a former IBM consultant and electronic engineering graduate, now writing about all things technology-related. He is particularly interested in how emerging technologies will affect enterprise in the future. After completing a Masters degree in Electronic Engineering at the University of York, he moved on to become an IT consultant at IBM UK. Gaining knowledge and experience of big data and its business applications, he specialized in the analysis and processing of sensitive data. Specific interests include big data analytics and strategy, natural language processing and machine learning.
Source:Digital Media Online. All Rights Reserved