Skip to content
Skip to main content
Author: Tony Hirst

A dark future for warehousing?

Updated Thursday, 14th March 2013
Dr Tony Hirst asks: If robots are at work, do we need to have the lights on?

This page was published over 9 years ago. Please be aware that due to the passage of time, the information provided on this page may be out of date or otherwise inaccurate, and any views or opinions expressed may no longer be relevant. Some technical elements such as audio-visual and interactive media may no longer work. For more detail, see how we deal with older content.

3D robot

Walk around the business district of any major city and you're likely to see architectural styles that show a preference for windows (the Shard, apparently, has 11,000 glass panels) - getting natural light into office spaces is seen as a Good Thing. In contrast, a walk around an industrial estate will reveal hectare after hectare of windowless rectangular boxes. If you can cut your lighting bill by £10,000 a year simply by changing the lighting, just imagine the cost savings if you could cut out the lights altogether.

While researching material for the revised robotics unit in the new Open University (OU) introductory course Technologies in Practice, I was reminded of the notion of "dark factories" which I'd first come across when developing the OU robotics course that inspired the new unit (Robotics and the Meaning of Life). The opening to a November 2001 article from The Engineer - And then there was darkness... - sets the scene as follows:

Every day the trucks turn up. They drop raw materials at the factory door. They pick up finished product at the other end. The factory operates 24 hours a day, seven days a week. There is no jovial banter, no 11 am coffee break, no demob-happy Friday afternoon. There are no windows and no lights. There is only darkness and the noise of robots working without rest.

As the article further points out: 'The main hurdle is stitching all the different technologies together. Integrated vision and inspection systems will also be required to make lights-out factories work. If no humans are involved in a manufacturing process, you must be sure that everything works together accurately. The more you stitch together, the more danger there is of the whole thing failing.'

But what if the process were a similar one, such as the movement of stock around a fulfilment warehouse? Perhaps the future is not one of dark factories, but dark warehouses?

If you've ever made a trip to the OU's Walton Hall campus in Milton Keynes (just half a mile or so away from the Red Bull Formula One motor racing team's factory), driving up the M1 from a southerly direction, you may have noticed a large Amazon fulfilment centre, just off Junction 13. A Daily Telegraph article - Behind the scenes at Amazon's Christmas warehouse - describes life there in the run up to Christmas 2012:

  • "As soon as products enter the warehouse, they are scanned and logged within 12 hours. The products are separated into different sizes, and then file along conveyor belts to one of 50 'receive stations'."
  • "[T]hey go straight to an area of the warehouse called 'mass land', which consists of pallet-loads of books and other high-volume items, such as Kindles".
  • "Shelvers pop things into a spare space, scan the item and then the code on the shelf. ... The only rule is that no two similar products can sit next to each other, to minimise human error while 'picking'."
  • Scanners group "orders together (about 80 items) based on when they have to be dispatched, and use algorithms to forge the most efficient route for the picker through the maze of numbered aisles and shelves."
  • "Once the pickers have collected their prescribed orders, they put the items into orange boxes, which travel via conveyor belts down to the packing area."

And so it goes... at least, for now. Because a year ago, Amazon bought Kiva Systems, a company that has developed a robotic warehouse picking systems that brings shelves to the the pickers, rather than requiring pickers to wend their way through miles of shelving each day.


So how might this impact on lighting? As the video clip from IEEE Spectrum shows (see also: Three Engineers, Hundreds of Robots, One Warehouse), the human picker still needs to be able to see to pick items off the mobile shelving; but do the robots need the lights to be on? Infra-red or ultrasound sensors to support collision detection means the robots don't need visible light to avoid bumping into each other (or us!) and while optical 2D bar codes do need visible light to be scanned by the robot's on-board cameras, this could be achieved by robot mounted torches. (While that would add a small power overhead to each robot's activities, LED lights are low power and long-lasting, so should require little maintenance once fitted. I also wondered about optical bar codes needing to be kept clean - I wonder if Kiva automated warehouses also have robotic floor cleaners?).

One step even closer to the notion of the dark warehouse - indeed, a warehouse where stock is largely inaccessible to human operators - means you can have enclosed areas within a warehouse that don't require lighting at all...

The Auto Store robot system developed by Scandinavian logistics firm Hatteland stores bins within a three-dimensional grid accessible only to robot pickers.

Three-dimensional grid-based storage also seems to scale up to bigger, palletised items, as the Power Automation Systems Powerstor "deep lane storage" system demonstrates:

As someone with an Artificial Intelligence (AI) background, I have to admit that the first thing I thought of when seeing these videos was how the machine knows how to route the movement of the bins or pallets efficiently. The problem reminded me of a three-dimensional version of a 2D sliding-8 puzzle, which has been widely used as a classic "toy problem" used in AI research. Solving that sort of problem efficiently is hard enough, but I'd hate to be the maintenance engineer if one of the trolleys came off its track!

Want to know some more? Check out:

  • Engineering: an active introduction (TXR120) - Residential School with a 1 day robotics activity.
  • Keep on trucking customer service - an OpenLearn introduction to supply chain logistics.
  • An episode of the OU co-produced BBC Radio 4 series "The Bottom Line" on logistics, from October 2010.

Become an OU student


Ratings & Comments

Share this free course

Copyright information

Skip Rate and Review

For further information, take a look at our frequently asked questions which may give you the support you need.

Have a question?