Five roles, from one machining line to eight plants abroad and at home

Mandate

Put seven plants and the overseas business on common systems, answer the shareholder's expectation of domestic core software and make the plants defensible.

Outcomes

• Moved 11 Chinese legal entities and six of seven plants to domestic ERP and MES software between 2023 and 2025; month-end close fell from eight working days to five.
• Brought 148 of 214 dealers onto a regionally hosted portal; warranty claims now settle in 16 days instead of 41 and first-pass parts fill rose from 81 to 92 per cent.
• Segmented eight plant networks for RMB 78 million after the 2023 Hefei incident; 52 vendors now use brokered, recorded sessions.
• Scaled three of eleven AI pilots and stopped six, with every result reported to the strategy committee.
• Cut China licence and support costs by RMB 41 million a year and custom programs from 1,180 to 270.

Decisions in context

I became CIO in March 2019, three months before Ruoshan's Hong Kong listing, when my predecessor retired. The president's brief was short: one set of systems for a company that had grown by adding plants, a credible answer to our largest shareholder on domestic software, and plants that could withstand the incidents other manufacturers were starting to report. I organised the department around seven leaders: applications, plant systems, infrastructure and cloud, operational-technology security, data and AI, international systems, and architecture with vendor management. Revenue has grown from RMB 24.9 billion in 2019 to RMB 38.6 billion in 2025, and the overseas share from 35 to 47 per cent, so most of the programme has run while the business it serves was changing shape. The two hardest moments, the failed Huai'an cutover and the Hefei stop, are described on my impact page.

Reason for the move

Current role. Not seeking an executive move; considering one independent directorship with the Ruoshan board's consent.

Mandate

Give the overseas business its own systems for parts, warranty, telematics and regional finance.

Outcomes

• Launched the first dealer parts-ordering portal with 40 dealers in Southeast Asia and the Gulf.
• Moved three regional companies onto the group ERP template, ending their separate local accounting packages.
• Built the warehouse system for the new Rotterdam parts hub, opened in 2017.
• Connected 19,000 exported machines to a telematics platform by the end of 2018.

Decisions in context

Ruoshan hired me in January 2016 to build what its overseas business lacked: systems that let a dealer in Jakarta or Dubai order a part, file a claim and see a machine's history without a chain of emails to Changzhou. I spent the first two years in Singapore, close to the dealers who were growing fastest, and the third in Changzhou as deputy CIO. My team of 48 launched a parts portal with 40 dealers, moved three regional companies onto the group ERP template and built the warehouse system for the Rotterdam parts hub when it opened in 2017. We also connected the first 19,000 exported machines to a telematics platform. The work taught me that a dealer network is a set of independent businesses, each with its own reasons to share or withhold information, and that a system offered to them succeeds only when it solves a problem they already feel.

Reason for the move

Appointed CIO when my predecessor retired.

Mandate

Turn a loss-making delivery unit for equipment manufacturers into one that finished projects on time.

Outcomes

• Delivered execution and planning systems at 31 equipment manufacturers.
• Raised the share of projects going live on the contracted date from 38 to 74 per cent.
• Moved the unit's delivery margin from a loss of 6 per cent to a profit of 14 per cent.
• Wrote the company's first cutover-readiness checklist, still in use when I left.

Decisions in context

Zhiluo hired me in May 2011 to run delivery for its equipment-manufacturing clients, a unit that was losing money and missing most of its go-live dates. Over four and a half years my team of 160 implemented execution and planning systems at 31 manufacturers of machinery, pumps, compressors and electrical equipment. The fix was mostly discipline: smaller scopes, written acceptance criteria and a readiness checklist that a client's plant manager had to sign before cutover. On-time go-lives rose from 38 to 74 per cent and the unit's margin from minus 6 to 14 per cent. I also learnt what a vendor cannot see from outside: how a client's plants really work, which of its managers will defend a project when it hurts, and how often a failed implementation is a failed process in disguise. I left to be accountable for what happened after go-live.

Reason for the move

Wanted to own the result of a system inside a manufacturer rather than hand it over at go-live.

Mandate

Introduce the parent's global manufacturing execution template in the Suzhou plant.

Outcomes

• Took the execution system live in welding and assembly in 2007 and in the paint shop in 2009.
• Documented 37 local deviations from the global template; the parent accepted 22 of them.
• Introduced first-pass yield measurement at 41 stations.
• Added serial traceability for hydraulic pumps and valves fitted to every machine.

Decisions in context

I joined Kellmann-Hua in September 2005 as lead engineer for its execution-system programme and became programme manager in 2008. The German parent wanted its global template in Suzhou without changes, and the plant wanted to keep working the way it always had. My job was to find where each side was right. We took the system live in welding and assembly in 2007 and in the paint shop in 2009, and I wrote up 37 places where the template assumed suppliers, shift patterns or inspections that Suzhou did not have. The parent accepted 22 of those deviations and asked us to change our own practice for the rest, which was usually the better outcome. In the evenings I studied part-time for a master's degree in software engineering, completed in 2010. I left in 2011 to see manufacturing software from the vendor's side.

Reason for the move

Wanted to see how manufacturing software is designed and delivered, not only configured.

Mandate

Keep PLC and CNC controls running on the cylinder-block, cylinder-head and crankshaft lines.

Outcomes

• Cut unplanned controls downtime on the cylinder-block line from 6.1 per cent to 2.3 per cent of scheduled hours.
• Introduced a register of program versions and of every outside party allowed to connect to the lines.
• Trained nine controls technicians, four of whom later led workshops.

Decisions in context

I joined Qingtan in July 1998, straight from Yanhu Institute of Technology, as an assistant engineer in the machining workshops of a state-owned engine maker in my home city. For the first three years I cleared PLC and CNC faults on night shifts. After the modem episode in 1999, when a machine builder's engineer loaded the wrong program from abroad, I started a register of program versions and of every outside party allowed to connect to our lines, which the workshops kept for years afterwards. In 2002 I became controls lead for three lines, with a team that grew to nine technicians, and we cut unplanned controls downtime on the cylinder-block line from 6.1 to 2.3 per cent of scheduled hours. I left in 2005 because I wanted to work on systems that ran a whole plant, and a joint venture in Suzhou was introducing one.

Reason for the move

Wanted to work on systems that ran a whole plant rather than a single line.