The debate over GlobalEye has mostly been about which aircraft Canada should buy. I keep coming back to what happens ten years after delivery, when the Royal Canadian Air Force (RCAF) wants to fit a new sensor. Will a Canadian engineering team have the records and permission to do the job?

Saab became Canada's preferred supplier in May, while making clear that it had neither a contract nor an order. The company offered to build, maintain and upgrade the aircraft with Canadian partners and to transfer knowledge and technology. That gives Ottawa something to negotiate. The contract needs to say exactly what Canadian firms will be able to do.

On September 25, Saab and the Defence Investment Agency signed a non-binding term sheet for negotiations covering six aircraft, a Canadian configuration and industrial participation. The agreement doesn't commit Canada to a purchase. Saab's announcement gives no detail on rights to the engineering data or authority over future design changes. Those subjects may well be on the table. They need to be settled before Ottawa signs.

CBC's September 9 report raised questions about radar performance, aerial refuelling, crew stations and GlobalEye's connection to NORAD. Saab has reportedly offered additional radars and a refuelling capability if Canada wants them. The RCAF needs to decide what it requires, with a clear account of what each addition would cost and how much development it would involve.

Adding a radar means checking its weight, power demands and cooling needs, then making sure it doesn't interfere with other equipment. A refuelling installation brings structural and fuel-system work, flight testing and revised crew procedures. More operator stations need space and power. An apparently small change can create work elsewhere in the aircraft.

The work leaves a trail of decisions, calculations and test results. A drawing may show where a component ended up without explaining why it was put there, or what limits the engineers worked within. Those details matter to whoever comes along next. So does knowing which version of the aircraft the approval covers, since maintenance instructions, training and spare parts have to match.

I think of these connected records as a modification ledger. It should let an engineer follow a change from the original requirement through the reasoning and tests to its approval, and identify the aircraft configuration it applies to. The engineers who made the decision won't always be there to explain it.

Consider that future sensor upgrade. With the right records, permissions and experience, a Canadian firm could assess the job and submit a credible bid. Without them, it may have to repeat analysis already paid for, or ask the original supplier to do the work. We could own the aircraft and still have very little choice about who gets the next contract.

DND already warns about this in its guidance on airworthiness support contracts. The minimum data access needed to meet airworthiness requirements, it says, may leave Canada unable to change where or how engineering work is done. Its advice is to deal with data and intellectual property access in the contract. Ottawa has identified the problem. This negotiation is a chance to act on it.

I've written in Vanguard about Canada buying sovereign hardware while renting the software. The same concern applies to engineering records. Canada needs permission to use them, update them and share them with authorized firms working on its behalf. That might require ownership in some cases and a lasting licence in others. What matters is whether the rights are sufficient for the work.

The CH-148 Cyclone is a reminder of how long a Canadian configuration can take to mature. Its contract was awarded in November 2004 and reset in June 2014 around a plan for five successive versions. The first Block 1 aircraft arrived in June 2015. Initial operational capability came in June 2018; Block 2.1 was cleared for operations in November 2021.

That history isn't a forecast for GlobalEye. It does show why Canadian modifications need their own schedule and approval process, with someone responsible for tracking every change to the design. Starting with an aircraft that already flies doesn't remove that work.

I would want the GlobalEye contract to settle four things.

Start with what Canada is buying. Establish a controlled design baseline, so everyone is working from an identified version and changes can be tracked against it. If capability arrives in stages, set out what each stage includes, how Canada will accept it and who is accountable. The manuals and training must match the aircraft that crews receive.

Then make the rights usable beyond the first support agreement. Specify which data Canada and its authorized contractors can use, modify or share after that agreement expires, and what technical support they will need. Identify restrictions imposed by other suppliers or export rules before they become a surprise. Access to a drawing still doesn't authorize a firm to approve a design change. That authority depends on Canada's separate airworthiness accreditation and oversight processes.

The records should arrive as the engineering happens. For each change, Canada needs the requirement, the reason for the decision and the supporting evidence, linked to the approval and resulting configuration. Digital files need consistent references and version histories, with checks that the agreed material has actually been delivered and is usable. Leaving engineers to reconstruct it all at handover would waste time and invite gaps.

Finally, give Canadian engineers defined jobs while the modifications are being designed. That is how teams learn why a decision was made and develop the experience to take on later work. Saab's partnership offer creates room for this. Ottawa's announced Industrial and Technological Benefits reforms, which encourage intellectual property transfers and investment in sovereign capabilities, should help secure those roles and the rights needed to carry them out.

Saab may be the right supplier. Staying with the original manufacturer for some work may also be the sensible choice. I want Canada to know which dependencies it is accepting, what they are likely to cost and whether it could choose another supplier later. A broad promise of technology transfer won't answer those questions.

Before signing, ask the negotiating team to walk through a future upgrade by a Canadian firm. Show where that firm gets the engineering record and what it is allowed to do with it. If the answer depends on goodwill or another negotiation, there is still work to do on the contract.