Panel session: Project Data Analytics, Hype or Reality. 23 Jan 2020
- Master of Ceremonies – Gareth Parkes, Construction digital and data leader, Sir Robert McAlpine
- Grant Findlay, Director at Sir Robert McAlpine
- Ian Lloyd, Director, Partner & Platform Strategy, Lloyd’s Bank
- Alex Budzier, Oxford Said Business School
- Grant Mills, Senior lecturer, UCL
- Steve Ashley, Digital Transformation Solution Centre Manager, Oil and Gas Technology Centre
- John McGlynn, Chair APM and Atkins
- Joanna Newman, PMI Board member and Vodafone
Apologies I didn’t capture everything; more highlights and interpretations than a transcript. My typing speed was incompatible with the pace of the conversation, but I hope that you find this blog useful nevertheless. We’ll try and get the video live soon.
Is every project really unique, which would reduce the utility of a data driven approach?
Alex – Thinking that every project is unique is very dangerous and means that there is a reluctance to learn from those who have gone before. There is a huge amount of similarity from one project to the next.
Ian Lloyd – We generally understand the human context of project delivery. We know that good projects have people who document their risks because they have thought about them.
Joanna – Projects are not unique, we are not inventing something completely new. We know that poorly documented projects tend to have a lower success rate.
Ian – I’m not advocating a highly documented project because this can stifle an Agile approach.
Grant F – The engineering challenges are similar. But the teams are different and variations in experience and management style are the key difference. Data can help to get inside of this. The data we inherited may not be the right answer, so we need to be cautious of applying analytics to everything.
John – There is a lot of commonality. What is changing is some of the techniques that we use. For example, modular construction. This will have the potential to make projects unique, but only until we have built up a body of experience with the new methods then the uniqueness becomes less. The challenge is to get the culture right, because some of these major projects begin to impact people’s lives.
Steve – When you really push people on it, projects aren’t that unique. There are classical project challenges. The best project managers are those who are effective at managing stakeholders. What comes out of this is the ability to free up time to focus on the people issues.
Grant F – Data trusts will enable organisations to securely pool data for the benefit of everyone. The trustees will have the final say on how data should be used and the extent to which it should be anonymised. The construction data trust is mobilising, with interest from over 70 organisations.
Steve – One of my roles is to encourage collaboration. Project analytics is one area where we can get organisations to work together, both within the sector and across sectors. We are working to mobilise a data trust within the oil and gas sector, working with Martin.
Joanna – We are at a tipping point. There will be an enormous amount of data coming from IOT, 5G networks etc. We need to find a way of leveraging this.
Academia – We see across a lot of clients and data. Different alliances deliver different amount of success. But as academia we can’t share and cross fertilise this data. We’ve seen a reluctance to share data but pooling data could be a solution.
Grant F highlighted an example where they had challenges with radiators rusting. The buildings were so insulated that the water didn’t move in the system and the dosing wasn’t protecting the system. When he looked he found multiple instances being managed separately across the business. He then found that there were 250 other insurance claims for where radiators were corroding. There was a massive industry issue that wasn’t being shared. There is a huge opportunity to use this data to change how we manage insurance too.
Alex recalled a discussion that he held in Australia. If suppliers are forced to publish their data at the end of projects it would force a change in culture. Organisations would win the project on the basis of innovation and their analytical ability, both of which are driven by data. It could be hugely disruptive.
Contracting and procurement
John – There is a long way to go on how we contract for project data analytics. How we award contracts. How we pool data. Atkins may have 10,000 projects yet how much data do we share internally; not very much. But this is changing.
Steve cautioned that if we insist on data being shared through regulation it tends to be sanitised. Pooling data makes the job a bit easier because the client or regulator doesn’t have a right to access the raw data to undertake analysis that the supplier may be uncomfortable with. It becomes a decision for the trustees. Pooling rather than sharing is a crucial difference.
Ian highlighted that they are moving beyond what is good for their organisation to what is best for the ecosystem. They then have an opportunity to innovate, do things differently and create new products.
Steve suggested that we need to look at the size of the prize. It’s about taking cost out of the system and focusing on specific use cases and aligning the data with answering those use cases.
Alex – We need to do something radical. Incremental change won’t get us to where we need to be. How do we deliver a x10 impact?
Joanna – The x10 challenge is about automating the transaction, introducing predictability. We have a repository of data about what works well and what doesn’t. We can then refine off the back of that knowledge.
John – Project managers need to understand how best to leverage data and advanced methods. We can then free up their time to do all the soft skills.
Ian – We need to understand the lineage of data and how reliable it is. If the data safe to use. How can I evidence it. How do I get people in the data chain to ensure that their data is reliable. We need to think carefully how we use the data.
Alex – The training data sets amplify the biases, so we need to ensure that we understand this issue much better. Construction is not about automated decision making, like credit decisions in a bank. It’s about augmented decision making. AI and robotics can reshape how we deliver work. But the challenge is, if we automate the lower grades, what is the entry point into the profession.
John – Ethics comes down to the human element, questioning whether I am now going to do something I am safe or qualified to do. It is a key element of the APM’s Chartership standard.
Ian – Google are getting to a situation where data is encrypted and you own the key. You can move your data to / from the Doctor to the hospital. It’s your encrypted data.
Question: How can we speed up this learning. How can we escalate how we do a good job?
Steve – Can we get to a place where we use data to get better at estimating. The proof points that there is something in this that gets real board level commitment.
Grant F – Procurement routes are not conducive to good performance; we contract with an organisation who can deliver it cheapest and/or quickest then get surprised when it is late or over budget. The whole business model is ripe for disruption.
Question: How does this improve profit?
Grant F – Reduce costs for clients and then we can build more schools, more hospitals. We need to think about the macroeconomic picture rather than just focusing on profitability. But if we remove error, improve safety and improve productivity then this has benefit for everyone. Increased profit should result from this.
Steve – There is a society element to this. We need to take cost out of the system.
John – Step 1 is to stop margin downturn. Productivity becomes before profit.
Alex – If we can smooth out cashflow, it has a massive implication for the industry. Although construction margins are small, the biggest issue is actually cashflow for major construction companies.
Ian – If we can use data analytics to drive margin, from reducing costs to delivering a better offering. It is much more about an ecosystem play. Sharing data is really important.
Joanna shared her experience that lessons learned should be focused around narrative, rather than looking through lists of thousands of statements. Data analytics can help to discover these insights at the right time.
Alex commented his belief that the London2020 learning is ‘utter ****’. It’s not focused on learning, it’s about self–promotion. It’s about focusing on the problem not the solution.
Ian – I am struck by the way that the Chinese solve problems. They reimagine the whole problem rather than incremental improvements through learning lessons. We are learning from the way that they approach the problem. Data can help to shape this.
Grant M – A lot of the structures should be there to learn not to apportion blame. Data can help to enable this.
Martin – I also provided my perspective, in that the lessons learned process is broken. It works for personal reflection but not as a means of organisational learning. The concept of everyone knowing everything about everything is flawed. We need to use technology to enable us to develop a heightened awareness of issues based upon the conditionality of the project,
Grant F – Technology is changing at such a pace that we need to refresh quicker.
Steve – We fundamentally need to do things differently. If we want to get to net –ero we need to do things differently. But we can’t do this if we can’t get the data. We also reduce the barriers to entry and let the small guy in to innovate and to learn from other industries. Huge behavioural changes are both required and also enabled.
Thanks to everyone who supported the event. Please spread the word and let’s try to change the world together.
Martin Paver is CEO/Founder of Projecting Success. In Dec 2017 he saw the opportunity to transform how projects could be delivered through the application of advanced data analytics and founded the London Project Data Analytics meetup, which has expanded throughout the UK and has grown to a community of ~4000 people. He delivers strategic and tactical solutions that integrate project management and leading edge data science and analytics.