He is proposing a network of 220km of 'SkyCycle' routes, mounted above existing railway lines.
Launching his plans, Sir Norman  said: "I believe that cities where you can walk or cycle, rather than drive, are more congenial places in which to live. To improve the quality of life for all in London and to encourage a new generation of cyclists, we have to make it safe." Thus far, I would agree wholeheartedly!
So you might expect CTC to be really excited that such a well-respected architect's firm as Norman Foster is putting forward visionary ideas for creating traffic-free cycling conditions. You might be even more surprised that the Daily Mail  portrayed CTC as merely carping about the possible steepness of the ramps leading up to these SkyCycle routes, and that it might be a bit windy when you get there!
We had in fact also raised some rather more fundamental reservations.
Firstly, accessibility. Unlike motorways, cycle route networks need to be easy to join and leave, if they are to be useful. Cycle journeys are local, mostly starting and ending at buildings on the local road network. Cyclists want to make these journeys following routes which are convenient and reasonably direct, as well as being safe and attractive. Dutch cycling design guidance recommends that cycle route networks should have a "mesh density" of a route roughly every 200m (which in practice means it should be about as dense as the road network), with a main cycle route roughly every 600m. Foster & Partners are talking about roughly 200 access points on a network of 220km of cycle routes – less than one access point per km.
Finding space to provide ramps even at this frequency will prove difficult enough though in a busy city. The sheer length of the ramps needed to get up to the height of a deck over an elevated railway line will inevitably make it hard to find space for these ramps, particularly if they are to link in conveniently into the rest of the cycle network using roads on the ground below.
However the really fatal flaw in the SkyCycle idea is its sheer cost. It is estimated that the first 4 miles alone of this network will cost £220m - and then there is another 130 miles still to go!
That is a huge sum to pay for not tackling the fundamental problem - namely the excessive volumes of motor traffic on our road and street networks. Surely the whole point of encouraging cycling should be to help reduce motor traffic. This is vital not just to create better cycling conditions, but to benefit everyone, however they choose to get around. Even the most car-dependent individuals would surely prefer to live in a world with less traffic. SkyCycle routes are merely an incredibly costly way of ducking that critical challenge.
There is an inevitable two-way relationship between cycle use and motor traffic. More cycling can help achieve the highly desirable aim of reducing motor traffic. But the opposite also applies: a willingness to reduce motor traffic is essential for creating safe and attractive walking and cycling conditions.
There may be specific locations where cycle routes following railway alignments could be extremely useful - and possibly cost-effective too. But in general terms, pie-in-the-SkyCycle dreams of elevated cycle routes are nothing more than an expensive escapist fantasy - escaping the need to reduce car use in order to create truly healthy and liveable streets and communities, not to mention a liveable planet for future generations.