Description – Challenge #1 

Keywords: Machine Occupancy, Process Planning, Process Simulation, Fuzzy Logic, Process Mining 

Category: Proof-of-Concept Project 

The Challenge: 
The production process of one of our partners in the continuous production is highly manually operated. The processing of raw material through the production line is planned weekly. Still, it often needs fast and flexible adaptions within the process. This creates high manual effort and leads to an efficiency loss.

The process includes 15,000 parameters and different up-and downstream variables that lead to multiple possible production processes across multiple machines. The industry partner aims to get an (AI-powered) process simulation and planning that facilitates the correct machine occupancy and allocation of the raw material mixes in the different machines.

The simulation shall recommend the best possible way through the production based on the weekly production plan. It should lead to a faster process, and a double-check with the feasibility of the production plan. The complexity of this challenge lies in the multiple parameters and the effects on them across the whole production line instead of just optimizing one machine. A good user interface should accompany the simulation for various stakeholders to plan work orders. 

Description – Challenge #2

Keywords: Sensor monitoring, camera-enabled sensoring, robotics, high-speed monitoring 

Category: Proof-of-Concept Project 

The Challenge: 
Our industry partner seeks to optimize a pick and place process. A vacuum robot arm supports this process by placing items from an infeed band to a blister band. 
Sometimes during this high-speed process (three items per second), the vacuum robot arm struggles to build up the required threshold, leading to a picker crash. The crash occurs when the robot is not perfectly matched to the items at hand. A sensor- or camera-based solution should give additional information on the placement of the items on the infeed band in order to direct the vacuum robot to the exact position. The visual surveillance can be applied either at the infeed band or blister band. This pairing of visual surveillance and robot shall lead to a reduced number of picker crashes.

Description – Challenge #3

Keywords: BOM transformation, data transformation, assembly tree

Category: Early-Stage Project 

The Challenge: 
One of our partners needs a BOM transformation solution in their ERP system. The Engineering BOMs (E-BOMs) as the master data are received from the customers’ PDM/ERP system. In many cases, the structure, classification and level of information are insufficient for other departments such as assembly or purchasing.

The industry partner would like to have a solution to transform the BOMs in the ERP system (necessary: SAP integration) based on the needs of the different departments. Ideally, you could develop a method for setting up the “ideal assembling-tree” (assembly BOM) based on a 3D functional model structure (BOM) exported from the CAD system. 

Description – Challenge #4

Keywords: Speech-control, Wearables, Operator Instructions, Workforce Management 

Category: Proof-of-Concept Project 

The Challenge: 

One industry partner wants to optimize the efficiency of operator instructions. On the shop floor, multiple tasks need to be worked on, i.e., material supply, handling of alarms, taking samples in due time and many more. 
Suppose an operator misses doing a sample or does not recognize that material is missing. In that case, this could lead to machine downtimes as the worst case. To ensure that the batches run smoothly, our industry partner currently offers a line management software that shall be enhanced by advanced and direct guidance of the operators. This can be done by speech-controlled instructions or device-enabled alarm signals (i.e., wearable tech or a speaker). Our industry partner is looking for a solution that enhances operator instructions and decreases problem-solving time using sound or visual notifications. 

Description – Challenge #5

Keywords: Centerline, Digital Machine Documentation, Setting Tracking, Gamificationon 

Category: Early-Stage Project 

The Challenge: 
One of our industry partners is looking for a digital solution for simplifying and motivating operator reporting to optimize the mechanical centerline process. Currently, mechanical settings are documented manually for few machine types before installation on the customers' side, with only some customers tracking changes after installation regularly (daily/ weekly).
On the one hand, some customers have centerline processes established (e.g., communication via Mail of significant changes in mech. adjustments to process owner) without having a central database. On the other hand,  others do not know altogether what kind of adjustment their machines have. Measuring, documentation, communication and analyzing the settings are based on manual work. The new solution should motivate and engage operators to track machine settings regularly. Additionally, a gamification approach to increase operator engagement in this particular use case is a goal.  

Description – Challenge #6

Keywords: AR/VR Trainings​, 3D Models, 3D Scans 

Category: Proof-of-Concept Project 

The Challenge: 
One of our partners is performing VR trainings for machine operations. Currently, there are no or wrong formats of 3D drawing from suppliers of machines or machine parts. This 3D data is beneficial and necessary to optimize customer training and provide easy-to-learn solutions for new operators, e.g., for machine safety training. Getting data - especially for existing machines from machine suppliers – is usually a long and painful task for their customers who want to apply the VR training. So far, existing solutions are too time and cost-consuming; an easy and cost-efficient solution for more significant rollouts is necessary. Thus, our partner is looking for an easy, cost-efficient solution that enables 3D scans of machines (GLB format) to upload in Microsoft Hololens 2. 

Description – Challenge #7

Keywords: Energy Management and Controlling, Anomaly Detection, Energy Forecasting

Category: Early-Stage Project

The Challenge:
 One of our partners is running an energy monitoring and controlling system for the whole factory (plant and production). They want to expand their energy management system with AI-powered recommendation to enhance energy efficiency while keeping up machine performance and output. A solution should include anomaly detection, energy forecasting based on historical and environmental data. So far, on the one hand, side, they created a low-energy machine (80kw) to monitor energy efficiency at the machine. On the other hand side, they have an energy management system for the factory itself. They want to bridge the gap including machine and factory data. The goal is to generate useful recommendations based on AI for the customer. 

Description – Challenge #8

Keywords: Pick & Place, Visual Inspection, Sensor Monitoring, Robotics

Category: Proof-of-Concept

The Challenge:
One of our industry partners wants to automate a picking process in their production. Currently, the production process of e.g. elastomer parts includes one process step in which different sizes and shapes of elastomer parts are mixed in one basket. From this step, they need to be selected based on shape and size, ideally, automatically.  
As of now, selected sites have introduced solutions providing partial automation, but these have not been applied at a larger scale due to a lack of performance in classification quality and speed. Randomly placed elastomer parts (e.g. O-rings) of different sizes must be automatically classified, taken out of a box and placed with a predefined position and orientation. The solution shall enable a visual inspection of the elastomer parts and a robot-picking on the right band. 

Description – Challenge #9

Keywords: Energy Harvesting, Small Sensors, Wireless Energy Source  

Category: Proof-of-Concept Project

The Challenge:  
One of our industry partners uses wireless sensors to monitor equipment during various processes. These sensors require batteries, which need to be changed regularly. However, due to the small size of the batteries, sampling frequencies are limited to conserve energy and prolong changing intervals. The industry partner is looking for a solution to enable wireless energy harvesting for sensors. As the sensors are mounted on pumps etc., energy to power the devices could be generated from mechanical vibration or temperature gradients. This energy could be harvested for supplying the sensor modules with further energy. Ideally, this would be a plug-in module that fits into the battery compartment of the sensor and manages the supply.  

Description – Challenge #10

Keywords: Scalability of ML, AI application, data transfer  

Category: Proof-of-Concept Project  

The Challenge: 
One of our industry partners introduced machine learning use cases in singular processes and machines already. They are now focusing on the scalability and transferability of ML models. There are different data sets and models for multiple applications in the field of injection moulding processes in place. However, they can not be transferred easily to new models. The goal is to re-use existing data and models for training the ML faster. The industry partner is looking for a solution provider which helps to create a structured approach for reutilizing data. This would enable the transfer to new way faster to develop more AI applications in the field of production. The focus area here is the injection moulding processes of elastomer parts.  

Description – Challenge #11

Keywords: Asset Monitoring and Settings, Data Acquisition, Digital Twin, Sensor-based Machine Monitoring 

Category: Proof-of-Concept Project  

The Challenge:  
One of our industry partners is using more and more recycled raw material. That changes in its quality and requires a constant adaption of the machine to keep up the quality. Currently, injection moulding processes have fixed operation point. High-quality demands require an adaptive injection moulding process to react to the varying quality of raw materials and ambient conditions. Injection moulding machines are closed systems, standardized interfaces do not offer all required parameters and the desired frequency.  
Hence, the key challenge is to extract process and machine data which is internally processed within an injection moulding machine but not accessible for external evaluation. Especially, the extraction of various continuous process data (e.g. pressure, temperature) at a high measurement frequency is necessary.  

Description – Challenge #12

Keywords: Visual Inspection, Quality Control, AI Inspection 

Category: Proof-of-Concept Project  

The Challenge:
One of our industry partners produces many different products on which the product logo is placed differently. In case of misplacement or damaged logo print, the product will be sorted out within the quality inspection process. This process is still manual. An AI-based camera system that indicates the quality of the logo placement on the products would be the solution. The quality (crispness, clarity, coverage) of the marking is critical as this is the brand image to the consumer on the retail shelve of the top retailers around the world. The ink pad printing process varies due to ink mix variations or humidity/moisture content of ink which makes the process highly complex and checking afterwards necessary. Ideally, the solution would have a camera vision system that inspects the product as it moves by on conveyors. 

Description – Challenge #13

Keywords: Sustainability, Risk Assessment, Supply Chain Transparency  

Category: Early-Stage-Project

The Challenge:
Due to the current supply bottlenecks, production planning is at risk. This creates bottlenecks that can hardly be absorbed by common production planning tools such as staff availability, plant capacity or shift utilization. A global shortage will therefore require not only price increases but also more transparency in the long term. It will be necessary to explain to the customer where raw materials come from, how they are obtained and transported, and whether the company's ethical principles have been considered in the process. The solution shall enable a continuous risk assessment across the supply chain based on sustainable KPIs. First, it shall enable transparency of sustainability analysis across the value chain and second it shall also enable a supplier ranking.  

Description – Challenge #14

Keywords: Quality control, Central management tool, FMEA 

Category: Proof-of-Concept Project  

The Challenge:
A new standard called IATF 16949 and automobile customers is requiring their suppliers to use the newest versions of FMEA, and Control-Plan (Flowchart). New FMEA (AIAG/VDA Harmonization) was introduced in 2019 and requires evaluation of any new processes in production according to this new FMEA analysis tool. At the moment, our industry partner is still using an old evaluation method for the quality control process FMEA which is done in excel. They are looking for a solution that helps to manage the FMEA process better since it is too complex to handle it with excel. Ideally, the solution enables data transfer from existing FMEAs, create an automized reporting function and can connect to the customer compliant system (FMEA tool). 

Any start-up with a technology/ product solving one of our industry partners challenges can participate in our 3-day remote Challenge

As a participating start-up, you commit to actively participate in the Start-up Challenge for three days.
You will create one of the two deliverables:

1. Concept to solve a challenge or
2. Create a prototype based on the data set you will receive. The scope of the result will be defined with the industry partner. The event is free of charge for you.

The application phase starts in calender week 17 – with the publication of the challenges on this page.
You are welcome to register here in advance so that you will receive a notification as soon as our doors are open.

First place receives €14,000 of the investment pool of all partners, second place €6,000 and third place €4,000.

On the last day of the Challenge, you pitch your solution and a jury, consisting of industry partners, will evaluate your pitch.
Based on this evaluation, the winners will be announced. 

During the whole Start-up Challenge you and your team will be supported by our mentors.

They are all experts in specific fields, e.g. data engineering, strategic designers, venture architects and dedicated industry experts. They will support you with their feedback and knowledge along the way. 

The Körber Start-up Challenge takes place31.08. - 02.09.2021

The start-up challenge is a remote event.

We provide the appropriate infrastructure so that you can develop and implement your POCs online.
Our mentors and facilitators will support throughout the event. 

You will use your own hardware and software to develop solutions during the Start-up Challenge.
Körber Business Area Digital does not provide notebooks, smartphones, or other hardware and/ or software. 
Headphones are recommended.

In addition to the development of the PoC, there will be opportunities for networking and sharing experiences.
We will also make sure to provide a decent amount of fun to keep energy levels high. 

Any OEM or manufacturer with a challenge to solve and is interested in working with start-ups can participate. Size, industry and location do not matter.

We have two streams of challenges:

1. You have a manufacturing efficiency problem and want to create a concept to solve it.
2. You have a clear project and need to find a technical solution to serve it.

As a participating industry partner, you commit to actively participate in the Start-up Challenge for three days, including your project team. Furthermore, you commit to an 8.000€ fee. More information can be found in the FAQ about the fee.

The participation fee is 8.000€.
It consists of 6.000€ for the investment pool (price money for the winning start-ups) and 2.000€ administration costs.

You can apply for the 2022 manufacturing efficiency challenge here.