Sequential temporal ordering and patterning are key features of natural signals, used by the brain to decode stimuli and perceive them as sensory objects. To explore how cortical neuronal activity underpins sequence discrimination, we developed a task in which mice distinguished between tactile ''word'' sequences constructed from distinct vibrations delivered to the whiskers, assembled in different orders. Animals licked to report the presence of the target sequence. Mice could respond to the earliest possible cues allowing discrimination, effectively solving the task as a ''detection of change'' problem, but enhanced their performance when responding later. Optogenetic inactivation showed that the somatosensory cortex was necessary for sequence discrimination. Two-photon imaging in layer 2/3 of the primary somatosensory ''barrel'' cortex (S1bf) revealed that, in well-trained animals, neurons had heterogeneous selectivity to multiple task variables including not just sensory input but also the animal's action decision and the trial outcome (presence or absence of the predicted reward). Many neurons were activated preceding goal-directed licking, thus reflecting the animal's learned action in response to the target sequence; these neurons were found as soon as mice learned to associate the rewarded sequence with licking. In contrast, learning evoked smaller changes in sensory response tuning: neurons responding to stimulus features were found in naive mice, and training did not generate neurons with enhanced temporal integration or categorical responses. Therefore, in S1bf, sequence learning results in neurons whose activity reflects the learned association between target sequence and licking rather than a refined representation of sensory features.

No abstract

In cognitive science, behaviour is often separated into two types. Reflexive control is habitual and immediate, whereas reflective is deliberative and time consuming. We examine the argument that Hierarchical Predictive Coding (HPC) can explain both types of behaviour as a continuum operating across a multi-layered network, removing the need for separate circuits in the brain. On this view, "fast" actions may be triggered using only the lower layers of the HPC schema, whereas more deliberative actions need higher layers. We demonstrate that HPC can distribute learning throughout its hierarchy, with higher layers called into use only as required.

How can you effectively manage the wells if you do not continuously know what they are producing? This is even more the case when the wells are being started-up for the first time. FieldWare Production Universe (PU) is Shell's real time well Virtual Flow Measurement (VFM) tool, which is running on 60% of Shell's global production and has enabled significant added value in the areas of real time surveillance and optimization. Production Universe has now been applied in the start-up of a number of Shell offshore projects in the Arabian Gulf and the Gulf of Mexico. The purpose of this paper is to describe PU added value that has been achieved in the following areas of these project start-ups under transient and steady state operations:"Provision of well flow estimates from the very start of commissioning for early indications of well/reservoir performance"Provision of well flow estimates from the very start of commissioning for more accurate hydrocarbon accounting/allocation;"Calculation of gas flaring volumes from the very start of commissioning, again for more accurate hydrocarbon accounting and GHG emissions;"Replacing wet gas flow meters on individual offshore wells with equivalent virtual flow meters, hence saving significant CAPEX and OPEX;"Replacing offshore bulk flow measurement for total platform gas and liquid exports with the sum of the aforementioned well virtual flow estimate, again saving significant CAPEX;"Control of corrosion/hydrate prevention chemical injection ppm based on well flow estimates resulting in enhanced pipeline integrity and chemicals OPEX savings Introduction and background Shell has had a number of large upstream, offshore projects recently in various stages of start-up and/or commissioning. These projects are critical in terms of large capital investment, reserves contribution and degree of difficulty (very deep water, subsea processing, long subsea multiphase pipelines, feeding large onshore gas plants etc.). Hence it is imperative to start-up the processes as quickly and efficiently as possible and yet maintain the highest possible standards of technical integrity, safety and environmental impact. Key aspects for efficient/effective project start-up are well/reservoir surveillance and hydrocarbon accounting - it is important to know how much the wells are producing and the composition of the fluid streams, for maximum production, flow assurance, asset technical integrity and accounting purposes. Ideally this would be achieved by using Multi-phase Flow Meters on each of the wells to physically and continuously measure the oil, gas and water flows, or by routing the wells to the Test Separator as they are progressively started-up. However, MFMs may not have been installed for all wells, and for wells that have them, they are usually not commissioned at the time of well start-up ?? MFM commissioning requires fluid samples from the wells - for subsea wells sampling is usually done robotically and at a later start-up stage. However, Virtual Flow Meters (VFMs) can be operational from the very start of production. Similarly, test separators may not be commissioned at the time of initial well production and if they are operational they are not suitable for tracking production from multiple wells. Hence, VFM is of significant value for well/reservoir surveillance and hydrocarbon accounting from the first instance of start-up, up to the time when MFMs are effectively commissioned and thereafter as effective insurance in case of individual meter failure. VFMs have also been used for the following:"replacing wet gas flow meters by continuously estimating well flows, incurring significant CAPEX and OPEX savings;"replacing offshore bulk flow measurement for total platform gas and liquid exports with the sum of the aforementioned well virtual flow estimates, again saving significant CAPEX and OPEX;"continuous calculation of gas flaring volumes from the very start of commissioning, again for more accurate hydrocarbon accounting (ref 8);"continuously estimating chemical injection ppm, hence correcting for well flow changes and safeguarding pipeline technical integrity by always ensuring the right dosage;"saving OPEX by minimizing chemicals injected. In most Shell EP operating companies a real-time software application known as FieldWare Production Universe (PU) is used for VFM, providing a continuous indication of oil, gas and water flow for all wells. PU is a data driven modeling application developed by Shell ?? the development background and operational experience within the Shell Group have been extensively described - see references 1, 2, 3, 4, 5, 6, 7. Using data driven models, PU provides a "virtual" three phase meter for all of the wells, all of the time. An overview description of PU follows.

Sequential temporal ordering and patterning are key features of natural signals used by the brain to decode stimuli and perceive them as sensory objects. To explore how cortical neuronal activity underpins sequence recognition, we developed a task in which mice distinguished between tactile 'words' constructed from distinct vibrations delivered to the whiskers, assembled in different orders. Animals licked to report the presence of the target sequence. Mice could respond to the earliest possible cues allowing discrimination, effectively solving the task as a 'detection of change'problem, but enhanced their performance when deliberating for longer. Optogenetic inactivation showed that both primary somatosensory 'barrel' cortex (S1bf) and secondary somatosensory cortex were necessary for sequence recognition. Twophoton imaging of calcium activity in S1bf layer 2/3 revealed that, in well-trained animals, neurons had heterogeneous selectivity to multiple task variables including not just sensory input but also the animal's action decision and the trial outcome (presence or absence of a predicted reward). A large proportion of neurons were activated preceding goal-directed licking, thus reflecting the animal's learnt response to the target sequence rather than the sequence itself; these neurons were found in S1bf as soon as mice learned to associate the rewarded sequence with licking. In contrast, learning evoked smaller changes in sensory responses: neurons responding to stimulus features were already found in naïve mice, and training did not generate neurons with enhanced temporal integration or categorical responses. Therefore, in S1bf sequence learning results in neurons whose activity reflects the learnt association between the target sequence and licking, rather than a refined representation of sensory features.