Created with Sketch.
Switch Lab
Choose Lab

Last couple of months 8 teams of bachelor students of IDE have been working on a design for things that predict the cities of things. The Delft Design Lab Cities of Things commissioned the assignment to explore design experiences with the use of predictive knowledge as a design material. The 8 teams did a great job in coming up with very different concepts. In this report I will introduce the projects, you can find the project videos and I will give a short reflection per project.

The kick-off of the project was 10 February with a kick-off presentation for the teams explaining the Cities of Things paradigm and the concept of predictive knowledge. See the slides below.

The assignment was described as follows:

Context: As a starting point for your connected ecosystem, select an everyday object in the city that already has connected capabilities, or that you know of to already be intelligent to some degree. If you prefer, you could also start with an object that doesn’t have any of these features yet, but you wish it had. Design: the behavior or the thing in specific situations.

Proposed steps to take:

(1) deconstruction of the intelligence of the things, (2) adding predictive knowledge to the behavior of the interplay, (3) building this into an engaging relation through the design of the interactions.

A good way to discover how our interactions with connected things are influenced by the integration of predictive knowledge, is to simply make designs for it. Through the design you make in this assignment, you provide more insights and knowledge about how we will use such products and what we think of them, so that is what we reach out for to you. As a starting point for your connected ecosystem, select an everyday object in the city that already has connected capabilities, or that you know of to already be intelligent to some degree. If you prefer, you could also start with an object that doesn’t have any of these features yet, but you wish it had.

The setup of the course is to come to a result via two or three lean cycles, also framed as MVPs. It turned out that in the first cycle the relative abstract assignment delivered a lot of ideas and sometimes confusion. In the following cycles, and after the first feedback round, the teams were able to find clear directions. Testing out the final MVP with citizens worked for all groups very well to understand what the essence of the concept was.

In the end presentation, the 8 projects were presented. See the videos below. Before diving into the separate projects is it is nice to share some general impressions.

The value of an ecosystem of objects – It was quite clear the value for the citizens, and the richness of a concept, is increasing heavenly with the number of objects connected. Different objects that are, not literally the number of objects per se. There is also a difference in objects that are part of the core ecosystem of use, and ‘objects’ that are meant for communicating information or controlling the core objects.

The iterations of design and testing are key – Even more than with other product designs, this type of connected intelligent products can only be designed with exploration testing. It is hard to imagine how the object will behave let alone how the user will interact with it. The MVPs-cycles structure worked very well in this project.

Intentional behavior delivers a strong value – Some of the teams had the objects taking an active role, products had an agenda, or at least took initiative in the interaction. That turns out a very strong stimulating factor for human-object interactions.

In the final presentation the 8 projects were divided into three categories: (1) Connecting people, (2) Exploring cites, and (3) Neighborhood services. I will introduce the videos in my own words and reflect on what I have learned from the project.

Connecting people

What role can connected objects play in connecting people in the city, and what will be changing when adding predictive knowledge to that objects?

Project Benchat

This group took the public bench as a starting point for their predicting system. The bench is responding on the person sitting on the bench and starts a conversation. The bench can capture the emotions of the person sitting, and is a matchmaker with other persons sitting on another bench in the city. With a lighting concept, the person is guided towards the other.

The project shows that a conversation started by an object can trigger interesting relations. I see also possibilities to add more city objects to the ecosystem and make the conversation even richer. It would also be very interesting to see what would happen if there are multiple connections made, and people are triggered into conversations while ‘commuting’ from one bench to another. It would be interesting to see how data from the behavior in the city and using the bench could be feeding the predictive knowledge.

Project Smart Community Garden

This project combined an already strong concept for community building in the city with the possibilities of the connected and predicted objects. The visitors can interact with plants and other objects in the garden in a natural way and at the same time, connections are made between visitors in the garden. This is even connected to things outside the garden. The community garden becomes a matching garden.

I like the potential of the community garden as it is logical to have multiple objects connected. Plants are also strong potential profiling objects by making a choice. This can feed the profiling machine. This concept has a lot of potentials.

I would be interested to see how it will work without the use of screens for interaction; finding real novice interactions, or using voice.

Project Playlix

This project was aimed at playing with children. The group did research into the way children would respond to the object. The central object in the playground is a playful intervention but also an input device for data on behavior and predictor of dangerous behavior.

The design of the object is strong and inspirational, it could easily work as a playful object. It is good that the object is not super defined from itself and open for children’s conversation what can be very useful if you like to understand the behavior.

It would even be more interesting to have more types of objects being part of the playground interacting with each other and the children, as a true actor-network. Now the two other types of objects are only data-reading and visualizing objects.

Exploring cities

Bringing connected objects in the city to life as part of a way to explore the city makes a lot of sense. Integrating multiple objects, and building interest profiles through the interaction and choices made by the human interacting with the city things.

Project Qimmy

In this concept the citizen can wander around the city with an audio guide that is part of the earbuds. The interesting part is that the guide is not a voice of an imaginary person, but is the collective conversations with the buildings and other points of interest. Which makes it a very rich way to explore the city.


There is room for even more interactions and building up knowledge of the wanderers by mixing the single information and the collective knowledge. Having mirrored trips based on earlier explorations in cities etc. There is a lot of possibilities with this concept.

Project Moov

Another way of exploring the city, this time the key object is a sharing scooter that is used by two friends and is leading the friends through the city exploring new places, doing little games, etc.

Predictive knowledge is easy an element of making these explorations in the city rich and connected to others. Having direct interaction with the choices and responses of the users would be even more interesting. I can imagine that this could be a service to leverage the shared scooter plans.

Neighborhood services

What functional uses can connected objects in the city have, especially for neighborhoods?

Project Hero 1

A delivery service that is autonomous driving around the city and is open for use by citizens, for peer-to-peer lending, not only for professional delivery. It is like a driving cupboard.

Adding peer-to-peer services is very interesting. Also thinking about connecting with other objects in the city or the objects that are moved around. Maybe the Hero cart can take an active role in connecting the citizens? Limiting the cart to a neighborhood can be adopted by the neighborhood is very interesting too; a community-owned last-mile delivery tool.

Project Kliek uit de Muur (leftovers from the wall)

A true community service; share your leftovers with others by using this ‘marketplace’ for leftovers via the central dispenser and smart food containers.

The predictive knowledge is now used mainly for convenience, to quickly recognize the meals and connect to the system. There is a lot of potentials to make the system more an actor on itself initiating certain cooking by supply-and-demand.

It is also nice that these type of systems could very well work without the central dispenser, might be a nice function of Hero 1…

Project Green for Good

What about waste management based on intelligent routing. Something that is a no-brainer and experimented before. And what if you can let the waste be recycled. What if you will steer consumption towards the necessary waste? That makes an interesting waste collecting system.

What if the waste bins at home are connected too? That would make it even more integrated. I like the potential of scaling this concept, it is strong if a project can grow from an initial basic working concept into more functions.

Follow-up research

What does it mean for a designer to use predictive knowledge as design material? I am interested in that question and the insights from these 8 teams will be part of the research I hope to complete later this year. I will combine this with insights from other student projects, earlier interviews with professional designers, and more.

I also think there is potential in some of the projects to combine to a neighborhood-focused service to build on the collaboration with intelligent things. I might consider combining it into a graduation research project. Keep you posted!

Lab Director
Iskander Smit