How advanced data analytics will transform project delivery - PowerPoint PPT Presentation

How advanced data analytics will transform project delivery Aberdeen Meetup Launch Event Martin Paver CEO / Founder www.projectingsuccess.co.uk martinpaver@projectingsuccess.co.uk +44 777 570 4044

Crossrail

What Happens to the Data?

Lessons Learned Systems Undertaking a portfolio of work (client projects) improves productivity and reduces costs and schedule uncertainty We take this….. And abstract it to this….. New technologies can have significant and unexpected impact on cost and schedule; include for this in contingencies

2019 Analysis “Since 2011 fewer that 25% of oil and gas projects have been delivered on time, projects averaging 10 months delay and coming in around 35% over budget” https://bit.ly/2T7yKnL

Why We Need A Different Approach Lessons Learning from Leveraging Learned Experience Experience We’ve been trying it for 40 years The purpose of machine learning Lesson are not learned just repeated Projects are still late and over budget But also implies human learning

Fundamentally? • What is the predisposition of the work to variance? • Can we predict it? • How do we test for it? • How do we treat it and change the future? Project DNA Evidence based, tempering against bias.

Siloed Data But we have siloed and unconnected data….

Data in Separate Tools or Pooled Data? Tool Driven Implementation strategy driven by tool selection. Primavera/ASTA, Risk Tool, BIM etc. Considerable tool integration challenge. Tool Driven Platform Driven A platform that integrates multiple tools. A one stop shop that integrates database and tools for a project management or BIM centred use case. Vendor lock in. Data Platform Driven Driven Data Driven Connected data is at the core of the solution. Tools and platforms are used to capture, ingest, process, visualise and provide insights.

Increasing use of Low Code O365 tools Data Robotic Process Automation Driven Expert Support for ‘the difficult stuff’

Data Trust: Architecture Data Trust: Architecture Project 1 Project n

Project Analytics Data Trust Enabling organisations to securely pool data for the benefit of the collective More in the pipeline

Process Automation

Removing Repetitive Processes 71 step process automated.

Removing Repetitive Processes Removing Repetitive Processes

Scheduling Real time update of assigned tasks Scheduling Corpus and Context Extract Triples WBS Elements EVM data Recommendations Resourcing Adaptive Scheduling Weather Supplier performance Dependencies Risks etc

Risks Risk Risk Informed risk budget trends registers A once through process Systemic Risk Connected risks Risk lifecycle Risks-Issues-Lessons Risk mitigations Leveraging Risk Experience

Stakeholder Management Static Analysis Or Adaptive, dynamic networks, reflecting real time feedback and historical performance of specific groups/individuals

Bid Analytics 3. Will increase significantly as bid PWin feedback is added. Aim – to use existing contract data to predict future bid outcomes Data - using ~10,000 contracts from construction industry 2. This increases where there is a 1. We can predict bid greater amount of unique data, outcomes to an such as from Network Rail (50%) accuracy of 30%. and North Tyneside (91%)

What Insights Can be Derived from Data • Kier is predicted to win the opportunity with 73%. • Mace are likely to come in 3rd place with a score of 68.6% Suppliers who bid for more contracts (size of bubble) tend to have a higher • We believe that the prediction can be overturned by: average score and lower • Improving BIM score by 6.7%, becoming top quartile. spread of scores • Increased focus on sustainability. • Improved networking with Brian Smith. Smaller organisations tend to have a lower average score and higher spread of scores *Bubble size = number of contracts bid for

A Critical T-Junction Ad hoc Data AI Future • We accept that our data is patchy • We believe in the vision • We acknowledge that its not a priority • We develop a roadmap to get there • We implement ad hoc improvements • We begin to lay the foundations • Data remains an exhaust plume • We upskill, attend hacks, reshape • Not really ‘invested’ • We are ‘invested’

We can’t change this alone

Mobilising a Force for Good

Developing a Uniting Vision: White Paper Examples of an integrated and united approach • Scotland data strategy • BIM But the data strategy for project delivery is either non existent or disjointed Why can’t we create something similar to leverage the opportunities within advanced project data analytics? Please join in.

Barriers to Adoption Its not on the corporate ‘to do’ list • Lack of a shared vision • Lack of evidence to support the vision • Understanding the investment case • Lack of skilled horsepower • Lack of data • Siloed • Poor quality

Barriers to Adoption • White Paper • Steering Boards • Meetups • Innovation • Apprenticeship • Hackathons • Robotic Process • Masterclasses Automation • Data Trusts

How Will You Engage With it? • This is progressing at pace in other sectors • Project delivery is a late adopter, but ripe for disruption • The capabilities are being demonstrated on a daily basis It isn’t hype • Some starting small, others more visionary • When it moves it will be difficult to catch up • Project management will be transformed

Contact Please find me on Linkedin : Also follow the Project Data Analytics Community Martin Paver CEO / Founder www.projectingsuccess.co.uk martinpaver@projectingsuccess.co.uk Martin Paver +44 777 570 4044

3,201 members 367 members 336 members

Community

Data Trust Paul Hamer, CEO of Sir Robert McAlpine wrote to 15 other CEOs about the creation of a construction datatrust. News A 10 partner innovation proposal was submitted on 30/10/19 to develop the construction data trust. Oil and Gas data trust is out for consultation.

AI Builder News

pyforest #pyforest -lazy is an opensource library that imports all popular hashtag#Python Data Science libraries so that they are always there when you need them. If you don't use a library, it won't be imported. News This is all done with a single line of code: from pyforest import * And if you use Jupyter or IPython, you can even skip this line because pyforest adds itself to the autostart. Installation: pip install pyforest Github repository: https://lnkd.in/f44-hSA https://www.linkedin.com/posts/parulpandeyindia_pyforest-python-ugcPost-6567249651899166721-JE1E/

Data driven decisions News News https://www.marketingtechnews.net/news/ 2019/jan/30/seodata-analyst-hybrid-why- you-should-advocate-data-driven-decision- making-content-driven-field/

News https://projectdataanalytics.uk/newsletter

Community Events

Community Events

Book now for #ProjectHack5 Community Events Book before the end of November to get a community loyalty discount! This is a community event and we need more sponsorship to make this event viable. Please speak to one of the team today if you can help.

Project Hack

Our Sponsors Our Sponsors Many thanks to our sponsors and supporters. The PDU Code for (2T in PDU Triangle) tonight’s event is: