The pandemic has forced companies to reconsider the role of the office in the life of employees and business in general. The concept of an office space has changed together with the approach to planning; generative design and artificial intelligence are of big help here.
Generative planning as a team member
Generative design was first used in machine building. It uses computer technology in the design process: the designer delegates some of the processes to the machine and this symbiosis allows for choosing an optimal project in accordance with the relevant criteria.
In respect to the office space it means that artificial intelligence can be “fed” any list of parameters that are important for planning a workspace. They can include the distance between desks — especially important amid the pandemic — as well as lighting, ventilation and heating settings. Using these parameters, artificial intelligence develops hundreds of seating and zoning designs that a human cannot create in such a short time.
It is important that computer algorithms suggest scenarios that we would have rejected straight away. The computer does not have blind spots or experience, so it fearlessly dismisses straight lines in design proving that the semicircle layout will save space and help to organize a seamless workflow.
For instance, one can make AI reach certain lighting parameters while reducing the number of power outlets and lamps. Or it can simulate the traffic in the office and arrange the desks so that they will not come in the way, let a group of employees pass by each other or encourage experts to meet more often and talk business.
Artificial intelligence is responsible for modeling and evaluating the initial variants, checking the efficiency of all possible solutions and creating a collection of the best ones. Naturally, humans have to take part in the process. An expert sets the parameters, adjusts them, analyses the best results, and then picks the best two or three from a human’s point of view. Then the variants are sent to the designer who adds details and their own vision. As a result, from a sole creator the human becomes the co-creator and even observer, while artificial intelligence becomes their partner.
Stages of generative planning:
1. The designer sets parameters for the system.
2. AI generates a field of options that meet the basic requirements.
3. The options are automatically compared to each other.
4. Top most effective solutions are picked up.
5. The designer examines the best options and adjusts the parameters for the machine to choose next steps.
6. The system suggests new options and ranks them.
7. The designer compares these options and checks the results.
8. The most preferred option is selected and then integrated into further work on project implementation.
How does it work in practice?
Generative design is being actively utilized in production and urban planning. For instance, the Dutch-based construction company Van Wijnen Groep has used generative design for its prefabricated housing project in Alkmaar, a city in the country’s province of North Holland. The team had to design affordable housing while conforming to local building standards with due account of the developer’s requirements. The company specialists developed seven parameters, which included architectural, financial and environmental requirements — such as the commitment to use three specific types of cottages as well as certain types of apartment buildings. AI generated the most relevant design options for a specific location, which helped to choose a project to suit everyone.
Generative design is yet to become a widespread trend in workspace design. But experts have already noted a massive interest in the technology from specialized professionals. Utilizing such design technologies requires a proper understanding of Building Information Modeling (BIM) — an intelligent 3D model-based process.
For algorithms, there is no big difference as regards the object of calculation and design, be it residential buildings, neighborhoods, factories or workspaces. Today, an international design technology company originally from California is recognized as a leader in the industry of generative design for workspace layout. The developer’s products include a generative design module that can produce a nearly infinite number of projects. The technology has already been used by architects to design Autodesk’s Toronto office. Upon receiving a system of given parameters and restrictions, AI managed to find, develop and test 10K layout options for three floors over the course of just a few days.
The parameters included the area of ceilings, walls and floors, the number of employed people, and lightning requirements. Employees’ preferences were considered as well — in particular, as regards their workspace neighbors, working habits and preferable interior, as well as staff interaction scenarios, noise level, workplace distractions, flexible space availability, game rooms, and a sufficient number of conference rooms.
As these thousands of design options contain data that indicate how well they correspond to each parameter, they can be compared and evaluated. This resulted in ten projects picked up for discussion. The work performed by AI allowed for choosing the most effective compromise solution for all parties involved in the office design project, despite the large number of instructions, preferences and expectations that had to be considered at the start of the project.
Advantages of using AI for office space planning:
• it takes on routine tasks, leaving us with the sole function of critical thinking;
• it can create designs that we could not imagine;
• it can calculate, offer and analyze all possible solutions;
• it saves time, finances and other resources (for instance, certain algorithms can reduce the amount of required construction materials);
• it allows us to provide arguments and explain the choice of a specific solution.
Generative design technology allows considering all relevant conditions for a project, including the given parameters of a building or site and the customer’s preferences. When working side-by-side with AI, designers and engineers enter key values into the system to be considered, such as design goals, construction materials, budget restrictions, and others. The computer calculates and visualizes the maximum number of possible projects, compares them, and selects the most efficient solutions.
No-one is saying that generative planning is better than human design. This is just another working tool that helps you quickly explore, test, and optimize your options and make a reasonable a balanced decision for any level of design complexity.
By Alexei Pazhiltsev, Head, Key Vendors Department, Syssoft