software + tech
Productivity Skilled labor shortages, which were a challenge before the pandemic, will potentially become even more acute according to models de- veloped by the Association of Builders and Contractors. More than 650,000 workers will be needed in addition to the usual quota to meet the demand for labour in 2022 alone. This is where knowledge management software, underpinned by AI, can offer real advantages, thanks to its ability to break down data silos and process information quickly and with extreme accuracy, enhancing resource productivity. Preparing information for search access is time-consuming. Traditional methods require investing significant amounts of human time and la - bour to interpret, label, and rank data, ready for inclusion in a search database. It is impossible to pre-empt every keyword and synonym that may be used and documents will require multiple labels to prepare for all eventualities that cannot be preempted. This leads to irrelevant returns when using traditional search methods. There has also been a huge increase in the use of email, video chat, messenger services, and other non-traditional channels for dispensing advice and information which is not readily accessible for collation and indexing. Capturing and reusing information conveyed via email, in- stant message or even video call transcripts is even more complex than with traditional documentation. Accessing knowledge created in MS Teams, for example, is challenging, especially since one meeting can cover multiple topics around a project. Adding to the challenge, all of these communication tools – iManage, email, MS Teams, Sharepoint – have different search interfaces, which require multiple and repeated searches to find information. With no mechanism for easy discovery, the majority of current retrieval systems rely on keyword searches, which enables users to combine words and modifiers to retrieve relevant information, but these often yield irrelevant results as the search parameters are so large and there is a lack of context. AI technology can overcome this by bringing all of these sources of in - formation together in one place, indexing and segmenting the knowledge created, and allowing it to be discoverable. Using AI-enabled technology allows a richer input of information as a query, including dragging and dropping full documents into the insight engine – which means results are more accurate, relevant, and refined. This is particularly useful for engineers working on projects where there are potentially thousands of different results per keyword. The context enabled by AI data discovery refines the returned results to a manageable number that can easily be reviewed and utilised to avoid a loss of productivity. New cloud-based software solutions, such as those offered by iKVA, can instantly be implemented and integrated with existing workflow systems to relieve the strain on resourcing issues. iKVA’s solutions, which harness AI, Advanced Machine Learning, and vector mapping technology, enable construction organisations to leverage the unstruc- tured data being generated from multiple sources to reduce wasted time and improve profitability. This avoids sourcing highly-specialised employees as well as costly upgrades to the legacy systems which are still frequently used across the industry.
Construction is one of the largest industries in the world, employ- ing around seven percent of the world’s working age population . At this scale, it is no surprise that the sector generates huge volumes of data on a daily basis; from field data captured by drone, to reports and contracts, data is continually being gathered from multiple sources but is frequently unstructured and difficult to access. Through necessity, contractors, engineers, and suppliers have pivoted to working and col- laborating digitally – using video calls to host meetings, site visits, and transact business, further increasing the amount of data being gener- ated. However, 96 percent of data from infrastructure projects is not used and 90 percent of engineering and construction industry data is unstructured. Big data – extremely large, unstructured datasets that can be analyzed by computers to reveal patterns and trends – is one of the most valuable commodities in the construction industry, enabling firms to improve cost effectiveness and operational efficiencies, while reducing business risk. In fact, analysis of big data increases a business’s chance of mak- ing better strategic decisions by 69 percent, so the desire to harness the power of big data is understandable – but the question is how? One of the biggest challenges facing today’s firms is accessing the data they already possess, to gain valuable business insights. The industry remains one of the least digitized sectors globally, partly because of the fragmented nature of its projects involving an array of contrac - tors and subcontractors. The problem of accessing siloed data points is exacerbated by the diverse computing systems used by construction firms. There is a clear opportunity for organisations to use Artificial Intelligence (AI) technology to process the unstructured data stored on legacy systems, collating it to enable users to discover knowledge from many sources across the organisation, even different geographi- cal locations. Commercial and procurement teams are routinely hampered by the lim- ited visibility of information and, additionally, lose a large percentage of their working day searching for the information required to perform their role effectively. With the industry facing increasing labour short- ages and the need to improve profitability and cost efficiencies, now is the time for companies to prioritise adopting AI solutions that can be used to discover data already in an organization’s possession. By harnessing the data that is already available, but currently undiscover- able, firms will be able to deliver improvements in the productivity, procurement, and planning associated with a construction project. Big data and AI: the building blocks of the future of engineering By Jon Horden
34
csengineermag.com
August 2022
Made with FlippingBook Annual report