Recipient Availability Profiles

AI Quick Summary

Failed deliveries are among the costliest last-mile inefficiencies — and become critical when autonomous systems replace human drivers who can improvise. This project introduces a recipient-facing scheduling tool where recipients indicate availability across times, locations, and authorised persons, feeding directly into route optimisation. The system skips addresses outside confirmed windows, holds packages when no slot exists, and offers fallback options (pickup points, paid guaranteed slots). It transforms delivery from a carrier-controlled push model into collaborative, recipient-driven scheduling — providing the demand-side intelligence that L4 vehicles and robotic agents require to operate without improvisation. Vision To make recipient availability a first-class input to logistics planning, maximising first-attempt success and laying the groundwork for autonomous delivery ecosystems. Project Justification Industry data shows 8–20% of parcels fail on the first attempt globally, with over a third of failures occurring because the recipient is not home. Each failed attempt roughly doubles per-parcel cost (Cordeau, J.-F. et al. 2024). Meanwhile, 67% of consumers report difficulty aligning delivery windows with their schedules. These problems intensify for autonomous delivery. A human courier facing a closed door can call the recipient, try a neighbour, or leave a package creatively. An L4 vehicle or delivery robot cannot — a failed stop means a return trip with no recourse. Planzer’s own L4 pilot in Bern means these are not hypothetical scenarios but gaps in an active deployment. Background Florio et al. (2018) introduced per-customer availability mappings to maximise expected successful deliveries through optimised route design, including planned revisits. Bonomi et al. (2025) extended this with recovery options (safe locations, collection points, redelivery), each carrying different operational costs. Burian et al. (2024) demonstrated that dynamically adjusting time-window lengths increases customer acceptance rates in both urban and rural settings. Planzer already offers a partial version including providing a safe drop location or narrowing the delivery time window for a fee. However, this is not a structured scheduling tool integrated into route planning, and the fees often discourage recipients from using the options. Proposed Solution Recipients interact with a scheduling interface — analogous to meeting-scheduling tools (see for example Let's Meet: https://letsmeet.ateleris.com/) — indicating availability across fine-grained time slots. They may specify multiple locations and authorised alternative recipients. This feeds into route planning to: •       Estimate delivery feasibility per address per day. •       Hold packages at the depot when no window exists. •       Signal drivers or autonomous agents to skip addresses outside confirmed times. •       Offer tiered options: free flexible scheduling, paid guaranteed slots, or self-pickup. Editability follows time-dependent logic: tomorrow’s slots are fully adjustable; today’s may signal a skip but cannot alter route plans. Alternative addresses are constrained once the package is sorted for a delivery zone. Pilot Approach A functional prototype tested via an extensive user study: participants interact with realistic delivery scenarios, entering availability, selecting fallback options, and completing structured questionnaires. Evaluation covers predicted first-attempt rates, route efficiency, satisfaction, and cost reduction potential — informing a subsequent operational pilot with Planzer.

Project Idea Metadata

• Project Idea Name: Recipient Availability Profiles
• Date: 4/2/2026 4:43:58 PM
• Administrators:

Project Idea Description

Vision

To make recipient availability a first-class input to logistics planning, maximising first-attempt success and laying the groundwork for autonomous delivery ecosystems.

Project Justification

Industry data shows 8–20% of parcels fail on the first attempt globally, with over a third of failures occurring because the recipient is not home. Each failed attempt roughly doubles per-parcel cost (Cordeau, J.-F. et al. 2024). Meanwhile, 67% of consumers report difficulty aligning delivery windows with their schedules.

These problems intensify for autonomous delivery. A human courier facing a closed door can call the recipient, try a neighbour, or leave a package creatively. An L4 vehicle or delivery robot cannot — a failed stop means a return trip with no recourse. Planzer’s own L4 pilot in Bern means these are not hypothetical scenarios but gaps in an active deployment.

Background

Florio et al. (2018) introduced per-customer availability mappings to maximise expected successful deliveries through optimised route design, including planned revisits. Bonomi et al. (2025) extended this with recovery options (safe locations, collection points, redelivery), each carrying different operational costs. Burian et al. (2024) demonstrated that dynamically adjusting time-window lengths increases customer acceptance rates in both urban and rural settings.

Planzer already offers a partial version including providing a safe drop location or narrowing the delivery time window for a fee. However, this is not a structured scheduling tool integrated into route planning, and the fees often discourage recipients from using the options.

Proposed Solution

Recipients interact with a scheduling interface — analogous to meeting-scheduling tools (see for example Let's Meet: https://letsmeet.ateleris.com/) — indicating availability across fine-grained time slots. They may specify multiple locations and authorised alternative recipients. This feeds into route planning to:

•       Estimate delivery feasibility per address per day.

•       Hold packages at the depot when no window exists.

•       Signal drivers or autonomous agents to skip addresses outside confirmed times.

•       Offer tiered options: free flexible scheduling, paid guaranteed slots, or self-pickup.

Editability follows time-dependent logic: tomorrow’s slots are fully adjustable; today’s may signal a skip but cannot alter route plans. Alternative addresses are constrained once the package is sorted for a delivery zone.

Pilot Approach

A functional prototype tested via an extensive user study: participants interact with realistic delivery scenarios, entering availability, selecting fallback options, and completing structured questionnaires. Evaluation covers predicted first-attempt rates, route efficiency, satisfaction, and cost reduction potential — informing a subsequent operational pilot with Planzer.

Failed deliveries are among the costliest last-mile inefficiencies — and become critical when autonomous systems replace human drivers who can improvise. This project introduces a recipient-facing scheduling tool where recipients indicate availability across times, locations, and authorised persons, feeding directly into route optimisation. The system skips addresses outside confirmed windows, holds packages when no slot exists, and offers fallback options (pickup points, paid guaranteed slots). It transforms delivery from a carrier-controlled push model into collaborative, recipient-driven scheduling — providing the demand-side intelligence that L4 vehicles and robotic agents require to operate without improvisation.