Deep learning for vision guided robotics Automatinglogistics · Deep learning for vision guided...
Transcript of Deep learning for vision guided robotics Automatinglogistics · Deep learning for vision guided...
Deep learning for vision guided robotics
Automating logistics
The team
Herbert ten HaveCEO
Kanter van DeurzenCTO
The team
Laurens van WijkCOO
Prof.Dr.Ir. Martijn WisseAdvisor
Logistics contains many heavy and repetitive tasks
Majority of warehouses (>80%) are manually operated
Labor is physically heavy and repetitive
Allowed margin of error is low
Main costs of operation:Warehousing: Order picking (>55%)
Postal sorting: Singulation (>65%)
Within the logistic sector variation is standard:
- Products vary- Packaging varies- Stacks vary
In order to automate these environments an effective and scalable solution for variation is needed.
Conventional robotics fail to handle product variation
Market opportunity
The global logistics robots market was valued at $3,060 million in 2015
Projection for 2021 is $10,518 millionCAGR of 25.4% from 2017 to 2021
Picking robots contributed to the highest share of this market
E-commerce is one of the biggest drivers. This will affect mainly warehousing and
postal services
Artificial intelligence
Picking
Through deep learning robots can learn to handle variation
Depalletizing
Quality control
We excel in determining Grasp Locations for Varying Objects through Deep Learning
Applications & markets
What?
Order pickingBin picking
Quality controlIdentification
For whom?
WarehousingPostal sector
Logistcs track record (a selection)
Warehousing logistics Parcel handling
End users
Integrators
Conversing with
[email protected]+31 6 53555955