Blog: On the path towards autonomous vehicles

On the path towards autonomous vehicles

-Click for reading part II-

Most new web services are deployed from a Mobile First philosophy.

Everything is connected, everything is mobile, but where is everything?

Put a note in your calendar to remember something, not when, but where you are to do the thing.

Wireless communication is perfect for moveable objects. Yet, during the 20th century we connected things upside down: A heavy TV was needing mains supply, but was receiving airborne TV-signals via antenna. When someone called you on the phone, they called the place of the phone; a living room or the kitchen, but not a specific person, as the phone cord was tying the device to the wall outlet.

Today huge (flat) TV’s are on cable while phones are cordless and mobile. However, the increase in wireless bandwidth even allows TV viewed on the move – streaming via the ubiquitous internet connection. We bring everything along.

But from where and to where?

Positioning via GPS will – for every one of us having a pair of eyes – be adequate even though the accuracy is often not better than 8-10 meter off the target coordinate. Such accuracy is not sufficient to determine whether you walk on the pavement or on the street, or if a car is occupying the right or left lane.

And further, if physical objects obstruct the signal, you are lost. For cars driving through high rise urban areas, in tunnels, or navigating between several decks in a parking house, this means loss of navigational capabilities; how to find out again?

Next step is self-driving vehicles

As devices now mostly know where they are, the next logical step is ensuring moveable objects will know where and how to go. For the automotive industry a set of definitions categorise a manually operated vehicle; a vehicle with few assisted functions such as taking the foot off during cruise control or hands off during parking assist; both foot and hands off while driving with aided lane keeping, or even eyes off once fully autonomous systems are mature. But roads, as we know them, are mixed zones with pedestrians, bicyclists, animals all behaving randomly (I once hit a boar on a German Autobahn at 2 am. My Volvo survived!). Hence the autobahn or motorway with its simple rules and no crossing objects (boars beware!) will most likely be the first place where autonomous cars are set free.

Certain autonomous applications are basically under centralised control such as unmanned metro trains or automated container handling at harbour terminals: people and other random encounters are kept entirely out of the premises.

For industrial purposes autonomous solutions appear promising to reduce operational cost: Extend the assembly or handling robot to let autonomous vehicles move parts to and from the assembly line. Move goods into, around and out of a storage. Discharge, sort and repackage from inbound containers to lorries delivering the last mile to customers.

In most cases one single and simple system of determining the position will not suffice. To make the system robust, different technologies will cooperate and recalibrate themselves by adding relative input (such as accelerometer and gyroscopic sensing of a vehicle) to e.g. a wireless positioning system.

Plenty of proven and operational automation technologies exist to track and steer equipment and goods. The challenge is selecting the optimal method to balance up-front investment and operational cost – while providing ample business opportunities to separate from competition and assuring the instalment is not leading to a dead end as technologies further mature.

-Click for reading part II-

More information:

Senior Advisor Jesper Meulengracht, jesper.meulengracht@glaze.dk, +45 70 23 50 05

We are currently offering the following positions in our Danish office