Inviting transparency in autonomous intralogistics

UX Design, Cartography

While we may be years away from having driverless cars, self-driving factory robots and forklifts are here and are already improving shop floor throughput. A large pitfall with the current state of internal autonomous logistics is that they are full of rigid and segregated processes, where each step needs to be controlled via various different software and multimodal touchpoints, thus adding significant friction to the day-to-day processes. Our client was working on a platform to unify these processes and reached out to us to help design the control system and interface.

Project Goal: Creating a system to manage intra-factory logistics operating autonomous forklifts.

Research: Field study and User interviews

To understand the context of the systems at hand, we visited a manufacturing unit of a major automotive manufacturer in Germany. Our team interviewed and shadowed the operators who managed the factory floors and logistic operations. We studied the tasks that they had to repeat quite frequently during their shift and the control systems they used for accomplishing them.

This study shed light on some pain points in the current practices and also helped pick out opportunity areas for interventions. The dynamics of the entire activity are complex and chaotic. It was at times surprising how under-developed internal logistic processes could bring a world-class production facility to a standstill.

Areas of focus and concept definition

We carried out a workshop to outline the key tasks that fleet operators tackle on a given day and grouped them appropriately. A priority analysis of these tasks in conjunction with the needs of the business helped us define the three most essential areas of the program that we would build in the first round of design.

The outcome of the workshops was a focus on three areas where we found the greatest opportunity for impact. They were: Orders overview, Vehicle Management, and Emergency actions.

Concept wireframing

Interactive Prototype

Excerpts

We continuously collaborated with factory operators and the engineering team of our clients to validate the design and its feasibility.

[ 1 ]

LEARNING: Control room operators often communicated with shop workers in case of vehicle or traffic issues. Since the factory floors were so expansive, the operators also needed to communicate the locations where attention was required.

IMPLEMENTATION: We had observed that the factory was divided into grids and the floor spaces were also visually marked with the respective codes. We integrated this grid system into our map.

Not just the map, there was opportunity to integrate this in other parts of the system too. The tablulated lists now included a secondary info row, that prompted the spatial location of the gates and the vehicles.

[ 2 ]

LEARNING: During emergency blocking, the operators needed to react quickly to avoid accidents. This meant blocking off areas on the shop floor and preventing vehicles from entering. But also in case of spills, the vehicles in that area too could be affected, and they had to be prevented from further spreading the oil. This was a process with dynamic flow, that depended on the individual scenario.

The solution to this was executed in two approaches. We did an A/B test to find out what was the most preferred method to carry out emergency blocking.

Separate approaches, and the preferred flow from testing

[ 3 ]

LEARNING: Exploring alternative visual styles with a goal to reduce eye strain for longer-term uses, we applied a darker-themed UI. While it succeeds in decreasing the amount of stress, we had given rise to a counter-issue - Halation.

The operators, who were generally of a relatively older age complained of the characters at times appearing when reading white text on a darker background. Though we made certain adaptations to alleviate that issue, the lighter interface came across a generally more comfortable to operate.

Adapting the map to the context

Designing the map in a way that helps the control operator and brings real value with respect to the task they are on needed us to make many cartographic decisions.

While we started off from a map that was focused towards presenting the spatial layout of the space, we felt that the map could be further augmented. That’s when began shaping the map towards a more schematic representation rather than just being a layout of the space.

This approach not only allowed us to further visually simplify the map keeping only the most necessary details, but it also created the opportunity to really enrich the map with the implementation of symbols towards making the map more dynamic and intelligent.

With a right culmination of the spatial layout with schematic cartography, we believe we’ve taken an approach that would truly aid the analysis process and help in making better decisions.

Gates

The gates are the spaces where the cargo would get picked up or dropped off. They are treated as live and dynamic entities that change their states through the course of operation.

States of the AGVs (Autonomous Guided Vehicles)

The AGVs are styled to indicate the state of the vehicles with appropriate symbol and color application.

Representing Areas

Using appropriate color and iconography that match the nature of the marked areas on the floor.

Text overlays and Rich Tooltips

Displaying text in conjunction with various elements on the map as tooltips to help better identify them.

Going Forward

These designs we showcased to partners and deployed at the testbed factories. It also helped the team extract and define interaction patterns and principles for the product as a whole. These patterns are helping drive the remaining blocks of the product such as monitoring, analytics, map editor.