Tech and construction: not so strange bedfellows
Wed 18 Dec 2019 | Gareth Parkes
On the face of it, there is little to unite the construction and tech sectors. Superficially at least, the construction sector provides a classic example of ‘waterfall management’: strict plans are formed at the outset of a project, and resources are fixed in place, even though the finished product might be three years away. It is a right-first-time, zero defects sector with detailed dependency scheduling. The tech sector, by contrast, is known for its not-afraid-to-fail, agile mindset, eschewing scheduling for creativity. Tech and construction, then, would certainly appear to be strange bedfellows
Times are changing. The proliferation of Building Information Management platforms on construction sites up and down the country neatly demonstrates that tech really can be embedded into the foundations of a building, and by extension, the construction sector. But how do we further embed the two, often conflicting, cultures?
To my mind, the answer is time and space: giving people a place to go, and the time to spend working on a new or different challenge. Such an approach is already being pioneered in the construction sector in the shape of the regular Project:Hack events. Now supported by a range of construction firms, using data from New York City construction projects, Highways England and Sir Robert McAlpine, these events started life 18 months ago from the germ of an idea to see whether it would be possible to apply data to construction. Added to this was some internal proof-of-concept data analytics work, demonstrating the power of the latest tools, technology and thinking, which ultimately revealed just how much the sector still could – and had – to learn.
The hackathons themselves grew out of the Project Data Analytics Meetup – in 2017 this was a 250-person strong community that came together in London once a month to think about how to better deliver large infrastructure projects to ensure that the next office, road, school or hospital is built using all the best available experience, understanding and know-how. This directly ties into the UK Government’s Construction Sector Deal, whereby projects must get 50% quicker and 33% cheaper, and indirectly into the Government Office of AI’s challenge to make the UK a global leader in data analytics, machine learning and AI. This community is now more than 4,000 strong, running regular events in five UK cities, with close ties to Australian groups and companies who are looking to get involved. It’s already gone beyond construction – the latest location to spring up is Aberdeen, with its links to the energy sector. The opportunities for cross-sector fertilisation and collaboration are huge – if you deliver projects, there’s something in it for you.
Soon after getting involved in the community, having heard talks from the likes of the Tour de France data team and other data thought leaders, Sir Robert McAlpine realised more could be done: it wanted not only to hear data stories but to write its own. Here, the team behind the meetups, Projecting Success, came up with their next winning idea: hackathons. Hackathons that give anyone the chance to play with real data from real industries, solving real challenges and helping the project delivery sector. Hackathons that support learning and development by running masterclasses from community experts, Microsoft MVPs, keynote speeches from tech and social value industry leaders, with prizes worth more than £1,000 for the winning teams. All set over a weekend at the Microsoft Reactor building in central London, with the cash from ticket sales going to cancer research.
Sir Robert McAlpine was happy to help: beginning by sharing some of its data, asking questions like “what health and safety insights can we get from our data?”. The response? Some interesting graphics and a barrage of feedback telling us how to improve both the data and challenges. Four months later, we were back for Project:Hack 2, armed with more data, better explanations and more explicit challenges, which included combing through data to unearth hitherto untapped insights.
With Project:Hack 3 came yet more data, more specific challenges and a fair bit of data wrangling beforehand to provide hackers with some training data sets – the 1st and 2nd placed teams both had supervised learning image recognition proofs-of-concept that found an object > 90% of the time.
At Project:Hack 4, 12 months after the first hackathon, there was a similar image recognition challenge, only this time without any training data, the task was to identify 10 objects. The working proof of concept is available online, with a tool that they can train in 20 mins to find any new construction object…and this team came third! The winning team created a working voice-activated safety observations tool, complete with an underlying database to categorise and prioritise safety hazards. Project:Hack 5 is next March, and the stakes will no doubt be raised.
For me, what these hackathons provide is not only the time and space for construction professionals to learn new skills and find new ways of working but, the starting point for a common language between the construction and technology sectors, which all too often can’t find the best ways to communicate.
This is a practical application of combining on-the-ground construction expertise with technology capability, which is helping both groups to improve together, by solving the important industry challenges. It’s a mutually beneficial system: the construction sector finds new ways of undertaking rapid R&D, bringing passionate, enthusiastic people into contact with organisations able to offer exciting careers in the built environment, whilst the tech sector finds a ready market for its innovations. If you want to change the world, come and build the future with us.
Tags:Big Data construction data analytics
Turning science fiction into science fact
Read More >>
Creative Destruction: Why the 2020 enterprise generation needs open platforms...
Read More >>
A cyber approach to coronavirus containment
Read More >>
The future of video streaming – Training AI to see with the human eye
Read More >>
The DCIM maturity model: An important step to data centre efficiency and adde...
Read More >>