When the demand for customised products at the cost of mass production increases, many companies are faced with very specific challenges. Still, the price tag of products is not the whole story. To remain competitive, companies must market new, innovative products and continuously optimise their production methods. We help them with all the necessary aid and expertise, as is shown by the following examples.
ALGORITHMS RESULT IN THE DEVELOPMENT OF AN INNOVATIVE BICYCLE COMPUTER AT ABSOLUTE CYCLING
Sensors in bicycle computers are very sensitive to a correct positioning and calibration. The installation procedure is therefore quite complex and, in practice, not really user-friendly. To make bicycle computers more accurate, Absolute Cycling wanted in first instance to study the use of sensors with an inertial measuring unit for improved GPS positioning. Secondly, they wanted to use these already available sensors to determine the pedalling frequency. Finally, they also searched a solution for measuring the travelled altimeters and gradient. The Dutch product development company started looking for a partner to tackle these challenges for designing an innovative bicycle computer.
Flanders Make and Absolute Cycling developed an accurate GPS positioning system based on a sensor fusion algorithm with data from inertial measuring unit sensors. They also developed an algorithm that allows to eliminate the use of a pedalling frequency sensor and instead makes use of the already available inertial measuring unit. In an ongoing follow-up project, the robustness and reliability of this algorithm are being optimised, particularly for cycling on poor-quality pavements or when uneven forces are exercised on the pedals. Flanders Make is also studying the feasibility of an algorithm for accurately measuring the travelled altimeters and gradient. Thanks to the already developed functionalities and comprehensive tests with a prototype specifically developed for this project, the new bicycle computer of Absolute Cycling should become more user-friendly and more accurate than what the current market has on offer.
“In an article on GPS positioning, Flanders Make emerged as a reliable partner with comprehensive expertise in sensor fusion, signal processing and the development of algorithms”, says Harm Giesen, CEO of Absolute Cycling. “They built on this expertise to tackle our challenges and, meanwhile, we’re already executing our second cooperation project.”
Absolute Cycling started in June 2019 with a crowd funding campaign through Indiegogo.
MORE ENERGY-EFFICIENT SYSTEMS FOR DUCO THANKS TO THE APPLICATION OF MACHINE LEARNING AND ADVANCED CONTROL
Since 1991, Duco is a leading European manufacturer of ventilation and sun protection systems. In response to the increasing demand for energy-efficient solutions, interaction with renewable energy sources and eco-friendliness, Duco launched its DucoBox Eco. This ventilation heat pump ensures the efficient supply of hot water, heating and fresh, healthy air in houses. Duco established contact with Flanders Make to examine to which extent its DucoBox Eco could respond even better to the comfort wishes of residents and to varying climate conditions.
The expertise of Flanders Make was used for implementing an advanced control strategy. To make sure that the capacities of the concept would be optimally used, the research centre developed an optimisation-based controller. The new algorithm makes use of models of the different Duco systems and of the actual residents and actual house. In this way, it can predict the impact of the different control actions on both comfort, cost and energy consumption. We also used Flanders Make’s expertise in machine learning to analyse the impact of the behaviour of the residents on the actual energy consumption.
Thanks to this study, Duco will be able to assess more accurately the energy and costs that a family will be able to save through the installation of a DucoBox Eco ventilation heat pump, either in an all-electric or in a hybrid set-up. Similar analyses and the newly developed control techniques will also be applied to future Duco systems as a stronger focus on renewable energy sources combined with a reduction of gas connections will generate complex energy issues.
A MORE EFFICIENT DESIGN PROCESS THANKS TO THE USE OF ADVANCED AND EFFICIENT MODELS FOR DESIGNING GOODYEAR CAR TYRES
When optimising the design of a car tyre, one must consider more than 50 criteria that the tyre must meet: behaviour during braking and driving, fuel consumption... All these performance criteria cannot be considered separately but must be optimised simultaneously. On top of this, vehicle manufacturers continuously want faster development times. Reliable, predictive simulation techniques therefore offer a distinct benefit compared to traditional time-consuming physical (experimental) tests. Goodyear Innovation Centre Luxemburg (GICL) established contact with FlandersMake@ KULeuven for developing an efficient, predictive numerical simulation method to calculate the dynamic forces between tyre and road surface and thus improve the efficiency of the design process for car tyres in terms of noise and vibrations.
Developing and validating new car tyre designs is a major challenge. Results from simulations can be used to assess the performances of a car tyre design. Because of the complexity and huge quantity of performance criteria, both advanced models and mathematical techniques must be applied to ultimately realise an efficient numerical design process. FlandersMake@KULeuven can boast comprehensive experience in advanced model reduction and was thus the optimum research partner.
A doctorate project was set up to develop the targeted tool, with the financial support of the Fonds National de la Recherche (FNR) Luxemburg.
The efficient, fully predictive numerical tool allowed to reduce the lead time of the car tyre design in terms of noise and vibrations from – initially – months to – currently – hours or days as it is no longer needed to build and test physical prototypes.
SIEMENS PLM SOFTWARE MAKES MAXIMUM USE OF DATASETS THANKS TO VIRTUAL MODELS
To develop new suspensions for vehicles or improve the life span of existing systems, it is crucial to be able to measure the forces on and stress in crucial suspension parts. Traditionally, sensors are used for this that are either expensive or difficult to mount correctly. Siemens PLM Software, a specialised software company, therefore performed research, together with Flanders Make, in ways to improve this process.
The goal was to use simpler and less expensive sensors that are also easier to mount. These sensors don’t measure forces or stress directly but link the measurement data with virtual models to enable estimating the targeted parameters through a so-called ‘virtual sensing’ approach. This significantly accelerates the validation process, thus considerably shortening the time-to-market of new suspensions.
Siemens PLM Software is currently considering the commercial integration of this approach in its Simcenter software portfolio and services.
Management and employees of Tenneco test cobots and digital work instructions in our MAKE LAB
Tenneco, a manufacturer of shock absorbing devices, wanted to examine the possibilities to integrate cobots (collaborative robots that are able to work together with people) and digital work instructions in their manufacturing line. For some time now, Flanders Make has a mobile living lab, the MAKE LAB, available for this. First, several types of cobots were tested to check their functionalities against the specific requirements. The selected model was then implemented in the MAKE LAB and complemented by a digital instruction system.
The functional demonstration set-up could literally be brought to Tenneco’s workplace so that both management and employees could try the new technologies.
“We were looking for a partner that could provide us with independent expertise and had the necessary infrastructure to bring this project to a successful conclusion", explain Stefan Peerbooms, Sr. Advanced Manufacturing Engineer, and James Richard, Sr. Advanced Manufacturing Technician with Tenneco. “The biggest challenge was to convince our entire organisation. Because everyone could participate in the demonstrations and could in an accessible manner familiarise themselves with the cobots and the new way of working, we definitely succeeded in this."
SUPPORT TO VAN HOOL FOR IMPLEMENTING FUNCTIONAL SAFETY STANDARDS FOR BUS AND COACH MANUFACTURERS
Bus and coach manufacturer Van Hool starts with the implementation of functional safety processes in their product design. They set up a partnership with Flanders Make, which has the necessary experience in the application of functional safety standards.
Using a case study within the ISO26262 process for bus and coach manufacturers, we go through the various stages in the process to implement the correct functional safety requirements. The partnership enables Van Hool to gain expertise and develop the necessary methods and tools to integrate the ISO standard into their current product design process in a transversal manner.
In the case study, the risks linked to passenger door functionalities were mapped. We defined the corresponding ASIL level (Automotive Safety Integrity Level) and, based on this, established the safety requirements that the system must definitely comply with in order to achieve this level. We also set up a framework for developing the necessary hardware and software. In this way, Van Hool acquires a complete survey of the main functional safety requirements for a bus or coach as well as the corresponding HARA analysis methods (Hazard Analysis and Risk Assessment), a hardware reliability assessment and an adapted and consistent software development method. Finally, we performed an ISO26262 gap analysis to identify further improvement actions.