Machine Learning (ML) is an incredibly powerful decision support tool typically used when an accurate prediction is required.
We are exploring clever ways to leverage the data produced on our Platform using ML scripts.
Huge volumes of data are being created throughout the design and construction process yet its embedded value is not being extracted and used.
This is further compounded by the siloed or non-collaborative way that each data creator like an architect or sub-contractor works.
We are developing a world-first “Bids That Learn” module to help project teams predict what capex and opex combinations lead to winning bids.
Ultimately our Platform will learn to predict the optimal combinations of mechanical, electrical, structural and façade items needed to achieve a performance goal over the buildings lifecycle – and keep it there.
When combined with the thermal engineering model, our machine learning algorithms will close the gap between design intent and performance.
The algorithms will provide evidence based decision making capabilities so that buildings are better designed to the needs of their customers’ performance goals.
The mainstay of our Platform is enabling project management teams to drive energy efficiency in their buildings.
But efficiency can be difficult to achieve - over-design wastes millions in unnecessary capital investment while under-design can result in a poorly performing building.
For the architects, engineers and builders faced with this challenge, contemporary techniques range from “guesstimates” and benchmarking to full thermal energy modelling.
The easy techniques are not accurate and the accurate techniques are not easy. This is why energy models are typically used only on larger projects as they take months to setup and skilled engineers to run.
At Benson Labs, we devise and implement faster whole-building modelling capabilities that are more flexible and simpler to use (no engineering degree required!)
We use procedurally generated models that require minimal set-up overhead. We deploy automated calibration scripts that improve model accuracy based on actual, real time performance. And best of all, our IP is based on EnergyPlus, a stable, modelling engine developed and supported by the US Department of Energy.
Our Platform and embedded modelling capability is fast, accurate and easy to use which means best-practice techniques can be applied on more buildings, by more people, globally.
We use blockchain technology to address the historical challenges around trust and collaboration in the industry.
Our labs are probing how the measurement and verification procedures (M&V) used to calculate a buildings performance target or guarantee can be reshaped by tokens, blockchains, smart contracts and tethered energy models.
The complexities to successfully execute an M&V plan are high, they go on for multiple years and often result in disputes between an energy service company and their client.
Our application will run on a custom-built blockchain, a powerful shared infrastructure that measures and calculates multiple unique and validated sets of building data that constitute the performance guarantee.
The M&V project will be governed by a smart contract, a replacement for the traditional M&V Plan that runs exactly as programmed without any possibility of downtime, censorship, fraud or third-party interference.
The blockchain will provide proof of provenance and ownership of the data that underpins an energy guarantee and do it in real-time. It will eliminate disputes and allow for adjustable baseline variances with ease.
Despite the known benefits to interoperability between software programs in the Construction industry, true interoperability is still lacking.
Data needs to flow seamlessly across teams, projects, and applications especially into the Building Information Modeling (BIM).
But there is a growing urgency to also integrate engineering models that can calculate and optimise the energy performance of a building. Currently this need is hampered by a lack of standards and frustrating workflows.
We are increasing interoperability between engineering models and the BIM model by creating APIs that enable the flow of data between our Platform and any other.
Leading BIM and CAD packages will be supported, removing many of the barriers and frustrations currently experienced by industry practitioners.
The right data in the right hands at the right time leads to stronger project outcomes and improves the way buildings are designed, built and managed.