Technology developments are making it possible to start automating the last-mile of parcel logistics. This raises the question of how these autonomous technologies will interact with humans. The goal of this project is to design a concept of a delivery robot for DHL, that maximizes acceptance by pedestrians.

Literature research and interviews showed that the appearance of the robot should be highly functional efficient to communicate what the function of the robot is and how it will behave. Cues that improve the understandability of the behavior should be emphasized in the design. Creating aesthetic associations between other logistics products and the robot makes it more familiar.

The level of in which the robot is perceived as a social emotional being should in be in line with peoples’ expectations and needs. How much anthropomorphization is applied on the robot, determines if the robot is seen as a machine, cute or creepy and forms expectations on how to use it.

The goal is to create a product which fulfills its function without disrupting regular pedestrian behavior and without eliciting negative reactions or sabotage from pedestrians. In order to achieve this, the interaction should be very low effort and intuitive. Pedestrians are not the customer and do not want to adapt their behavior dramatically to cope with tens of delivery robots on their sidewalk stroll. Intuitive interactions can be achieved by using the same way of interacting as with known entities, like other pedestrians or vehicles. Tests show that a design in which the behavior of the robot is modeled after standardized pedestrian behavior, could result in favorable interactions. In demanding situations, the robot will use a car-like blinking light, to communicate what its intentions are.

In user experience tests in virtual reality, small body cues from the robot were seen as an important communication tool on top of a predictable maneuvering. In the final design, the four wheels are made highly visible by a higher body and lights under the chassis. Suspension control allows the body of the robot to lean into corners and forward just before braking, to visually accentuate and communicate the impending action.

The sum of all the small cues and the path of the robot are intuitively understood by pedestrians and don’t need much attention while interacting. Pedestrians will feel in control, because the robot behaves predictable and it mingles in with the natural flow of pedestrians.

The behavior of pedestrians and robot are simulated based on the “social-forces model”. The outcome of the simulation is a first definition of desirable robot behavior and what parameters are needed to achieve this. It is found that the robot should have a less dynamic way of moving than pedestrians. It should steer slower and move with less speed changes. It should be very early in communicating its directions by already turning into that direction a few meters before encountering a pedestrian.

Although the social forces model is now used within simulations, it can also be implemented as the basic logic of DHL’s future delivery robot. However, more tests are required with live test subjects, to find definitive conclusions.

This master thesis contains the development of a product solution for wheelchair users. Based on behavioral science theories, persuasive technology principles and an analysis of the context and problems of wheelchair users, a strategy and product have been developed to help wheelchair user improve their wheelchair behavior.

Parts of the Southwest Palace of Sennacherib (700 BC), which were located in Nineveh (Iraq), are now destroyed by the ongoing conflict in Iraq and Syria. The palace rooms used to be decorated with numerous Assyrian reliefs. Luckily, digital pictures of the site are available thanks to an Italian expedition in 2002. The goal of this research is to physically reproduce the destroyed reliefs using this photo database. First photogrammetry was used to rebuild
the global dimensions of the reliefs. The details are reconstructed by fusing a highly detailed depth map retrieved with Shape from Luminance (SFL), together with a coarser Artificial Reconstruction. The fused reconstruction was evaluated with the eigenvalues of the Hessian. The eigenvalues give an intuitive measure about the shapes of the reliefs. The RMSE of the results improve up to 44.4% compared to the usage of an intensity (grey) image. The final 3D models are used to produce full size reproductions with 3D printing and CNC milling.