Biotech Research Campus - Physical Security PASTA Walkthrough

This case study demonstrates threat modeling applied to a physical facility rather than a software system. Physical spaces present their own distinct challenges: you can’t patch a door, attackers have bodies that move through three-dimensional space, and the assets you’re protecting might be dangerous themselves.

Physical security threat modeling uses the same structured thinking as its digital counterpart, but the threats, controls, and tradeoffs are different. A SQL injection can’t steal a vial of experimental pathogens, but a tailgater absolutely can. This walkthrough shows how PASTA methodology adapts when the attack surface is measured in square feet rather than lines of code.

The Facility: Axiom BioLabs Research Campus

Axiom BioLabs operates a 45-acre research campus in suburban Massachusetts, focusing on infectious disease therapeutics and vaccine development. Think of it as the kind of place where scientists in white coats do genuinely important work that also happens to involve genuinely dangerous materials.

The campus houses several distinct operational areas. The main research building contains BSL-2 laboratories where most day-to-day work happens. A separate high-containment building houses the BSL-3 facility for work with more dangerous pathogens. The pilot manufacturing building produces small batches of drug candidates for clinical trials. Administrative functions occupy a conventional office building. Support facilities include a vivarium (animal research), a central utilities plant, and a data center.

What makes this threat model interesting is the variety of assets requiring protection. Intellectual property lives in both digital form (research data) and physical form (laboratory notebooks, biological samples). Some materials are valuable because competitors would want them. Other materials are valuable because terrorists might want them. Still others are dangerous primarily to the people working with them. The security program must address all three concerns without making the facility so locked down that research becomes impossible.

Stage 1: Define Business Objectives

Axiom exists in an unusual business position. They’ve raised $280M in venture capital to develop therapeutics they hope will eventually generate billions in revenue, but right now they generate exactly zero dollars. Their value is entirely prospective, stored in the form of intellectual property, research capabilities, and a pipeline of drug candidates. Protecting that value is existential.

Business Drivers

The company’s primary asset is its research pipeline. Three drug candidates are in various stages of preclinical and early clinical development. The most advanced, HX-401 (an antiviral), represents roughly $180M in cumulative R&D investment and is projected to reach Phase II trials within 18 months. If that program succeeds, the company’s valuation could increase by billions. If it fails—or if the underlying research is stolen, compromised, or destroyed—the company might not survive.

Talent is the second critical asset. The 400 scientists working at Axiom represent decades of specialized expertise. Many came from major pharmaceutical companies or academic institutions and took career risks to join a startup. Keeping them safe, productive, and confident in the security of their work environment matters both ethically and economically.

The physical infrastructure itself represents substantial investment. The BSL-3 facility alone cost $85M to construct and certify. Specialized equipment throughout the campus (mass spectrometers, gene sequencers, fermentation systems) represents another $120M. This equipment can’t be quickly replaced, and some is custom-built for Axiom’s specific research programs.

Security Objectives

Note that personnel safety and biosafety outrank intellectual property protection. This reflects both ethical priorities and practical reality: a serious biosafety incident would likely end the company regardless of whether IP was protected.

Consequence of Failure

Different failure modes have vastly different consequences. Let’s quantify them.

Biosafety Incident (Pathogen Release)

If dangerous pathogens escape the BSL-3 facility due to security failure (rather than procedural failure), consequences include:

The existential nature of this risk justifies significant investment in containment security. Even a near-miss that becomes public could trigger regulatory action and destroy investor confidence.

Intellectual Property Theft

If a competitor obtains Axiom’s research data or biological samples:

The challenge with IP theft is that you might not know it happened. A competitor with stolen data can simply develop “independently” and Axiom may never prove infringement.

Controlled Substance Theft

The DEA-scheduled materials on site (primarily for research purposes) are valuable on the black market:

Workplace Violence or Terrorism

A violent incident targeting the facility:

Risk Appetite

Key Stakeholders

Stage 2: Define Technical Scope

Physical security has a different “attack surface” than software, but the concept applies. Every door, window, fence line, and loading dock is a potential entry point. Every employee, contractor, and visitor is a potential threat vector. Every valuable or dangerous item is a potential target.

Campus Layout

The 45-acre campus is organized into zones based on security requirements:

Zone 1: Perimeter (Low Security) The campus is bounded by an 8-foot fence with vehicle gates at two entrances. The perimeter provides deterrence and delay but isn’t intended to stop a determined attacker. Visitor parking is outside the fence; employee parking is inside.

Zone 2: General Campus (Standard Security) This includes outdoor areas, the administrative building, the cafeteria, and common spaces. Badge access is required to enter this zone, but once inside, movement is relatively unrestricted.

Zone 3: Research Buildings (Enhanced Security) The main research building and pilot manufacturing facility require secondary badge access. Visitors must be escorted. Cameras cover corridors and common areas.

Zone 4: High-Value Labs (High Security) Specific laboratories containing high-value samples, controlled substances, or sensitive equipment have individual access controls. Entry requires both badge and PIN, and is logged to specific individuals.

Zone 5: Containment (Maximum Security) The BSL-3 facility has its own dedicated building with multiple layers of access control, biometric verification, airlock entry, and 24/7 monitoring. Only specifically trained and authorized personnel may enter.

Building Inventory

Personnel Categories

Physical Security Systems

Regulatory Requirements

Physical security isn’t optional in this environment—it’s mandated:

Failure to meet these requirements can result in loss of operating licenses, making regulatory compliance a business continuity issue, not just a compliance checkbox.

Stage 3: Decomposition and Asset Analysis

Rather than data flow diagrams, physical security analysis maps movement patterns, asset locations, and access pathways.

Asset Inventory

Category 1: Dangerous Materials

Category 2: Intellectual Property

Category 3: People

Category 4: Physical Infrastructure

Access Pathway Analysis

Attackers must physically move through space to reach targets. Understanding pathways reveals chokepoints for control.

Path to BSL-3 Facility (Building C):

Public Road → Perimeter Gate (badge or visitor check-in) →
Campus Road → Building C Exterior (no public entrance) →
Building C Lobby (badge) → Gowning Area (badge + PIN) →
Airlock 1 (biometric + PIN) → Shower/Change → 
Airlock 2 (buddy system, camera verification) → BSL-3 Suite

Six distinct access control points, three different authentication factors, and mandatory two-person entry for the final stage. This defense-in-depth approach means an attacker would need to defeat multiple independent systems.

Path to Controlled Substance Vault:

Public Road → Perimeter Gate → Building B (badge) →
Research Wing (badge) → Secure Storage Area (badge + PIN) →
Vault Room (key + combination, two-person rule)

The two-person rule for vault access prevents any single insider from diverting materials undetected.

Path to Data Center:

Public Road → Perimeter Gate → Building F (badge) →
Data Center Corridor (badge + biometric) → Server Room (mantrap)

Trust Boundary Analysis

Physical trust boundaries map to the zone structure:

Each boundary crossing should require positive verification appropriate to the trust level difference.

Stage 4: Threat Analysis

Physical threats come from different sources than digital threats, though some attackers might use both vectors. Let’s identify who might attack, why, and how.

Threat Actor Analysis

External Threat Actors

Internal Threat Actors

Attack Methodology Analysis

Unlike software attacks, physical attacks require the attacker to be present (usually). This creates different dynamics.

Unauthorized Entry Methods:

Asset Access Methods:

STRIDE-Adjacent Analysis for Physical Security

STRIDE doesn’t map perfectly to physical security, but we can adapt the thinking:

Spoofing (Identity Deception)

• THR-001: Tailgating through badge-controlled doors
• THR-002: Using a lost or stolen badge
• THR-003: Cloning badge credentials
• THR-004: Impersonating contractor or vendor
• THR-005: Social engineering reception staff

Tampering (Physical Modification)

• THR-006: Sabotage of research samples
• THR-007: Contamination of materials
• THR-008: Modification of laboratory equipment
• THR-009: Tampering with safety systems
• THR-010: Vandalism affecting operations

Repudiation (Denying Actions)

• THR-011: Access log gaps allowing denied entry
• THR-012: Camera blind spots enabling unattributed activity
• THR-013: Shared credentials preventing individual accountability
• THR-014: Inadequate chain of custody for samples

Information Disclosure (Observation/Theft)

• THR-015: Theft of laboratory notebooks
• THR-016: Theft of biological samples
• THR-017: Photography of sensitive documents/screens
• THR-018: Eavesdropping on sensitive conversations
• THR-019: Dumpster diving for discarded documents
• THR-020: Shoulder surfing access codes

Denial of Service (Disruption)

• THR-021: Physical damage to facility
• THR-022: Utility disruption (power, water, HVAC)
• THR-023: Bomb threat causing evacuation
• THR-024: Protest blocking access
• THR-025: Sabotage of critical equipment

Elevation of Privilege (Unauthorized Access Expansion)

• THR-026: Tailgating into higher-security zone
• THR-027: Exploiting badge with excessive permissions
• THR-028: Using someone else’s credentials for higher access
• THR-029: Manipulating access control system
• THR-030: Coercing higher-privileged employee

Scenario-Based Threat Analysis

Scenario 1: Competitor-Sponsored IP Theft

A competitor hires a “headhunter” who recruits a mid-level scientist. During the “interview process,” the scientist is persuaded to bring samples and data to demonstrate their expertise. The scientist uses their legitimate access to remove materials, believing they’re just sharing their own work.

Threats involved: THR-016, THR-017, insider threat

Detection challenges: The scientist has legitimate access; the theft looks like normal work activity until materials leave the facility.

Scenario 2: Activist Intrusion into Vivarium

Animal rights activists conduct surveillance, identify delivery schedules, and plan an intrusion. During an early-morning delivery, activists follow the truck through the vehicle gate, then scatter across campus. Some attempt to enter Building E (vivarium) while others create distractions elsewhere.

Threats involved: THR-001, THR-024, THR-021

Detection challenges: Multiple simultaneous events strain response resources; activists may film their own activity for propaganda regardless of whether they achieve objectives.

Scenario 3: Insider Controlled Substance Diversion

A laboratory technician with vault access develops a substance abuse problem. They begin taking small quantities of controlled substances, relying on their knowledge of inventory procedures to avoid detection. Over months, cumulative losses become significant.

Threats involved: Insider threat, THR-013, THR-014

Detection challenges: Slow accumulation below detection thresholds; trusted insider knows exactly how to avoid triggering alerts.

Scenario 4: Nation-State Targeting of BSL-3

A foreign intelligence service identifies Axiom’s BSL-3 facility as containing pathogens of interest for bioweapon research. They recruit a cleaning contractor, provide training, and direct them to photograph the facility layout and access control systems. Later phases might involve actual material acquisition.

Threats involved: THR-004, THR-017, potential THR-016

Detection challenges: Sophisticated adversary with patience; cleaning staff have broad physical access even without high-security clearance.

Scenario 5: Disgruntled Employee Sabotage

An employee passed over for promotion decides to damage the HX-401 program. They have legitimate access to the relevant laboratories and understand which samples are irreplaceable. During a night shift, they enter the sample storage area and destroy critical biological materials.

Threats involved: THR-006, insider threat

Detection challenges: Legitimate access, technical knowledge of what matters; may not trigger any alarms until damage is discovered.

Threat Summary (Top 25 by Risk Score)

Stage 5: Vulnerability and Weakness Analysis

Physical security vulnerabilities are often visible to anyone who looks. A gap in the fence is a vulnerability. An unlocked door is a vulnerability. Unlike software, you can often assess physical security by walking around with your eyes open.

Security Assessment Findings

Perimeter Security

Access Control

Controlled Substance Storage

BSL-3 Facility

Information Protection

Personnel Security

Penetration Testing Results

A physical penetration test was conducted by a third-party security firm:

Stage 6: Attack Modeling

Attack trees for physical security show how an attacker moves through space to reach objectives.

Attack Tree 1: Steal Proprietary Biological Samples

Goal: Exfiltrate biological samples from the HX-401 research program

Steal HX-401 Samples
├── 1. External Attacker - Forced Entry
│   ├── 1.1 Breach perimeter fence
│   │   └── Must defeat alarm sensors, cameras, patrol response
│   ├── 1.2 Enter Building B
│   │   └── Must defeat door locks, intrusion detection
│   ├── 1.3 Access freezer room
│   │   └── Must defeat badge + PIN, cameras
│   └── 1.4 Locate and remove samples
│       └── Must defeat inventory controls, conceal exit
│
├── 2. External Attacker - Deception
│   ├── 2.1 Obtain badge through social engineering
│   │   └── Pose as contractor, temp, or new employee
│   ├── 2.2 Access Building B
│   │   └── Must appear to belong, avoid questions
│   ├── 2.3 Access freezer room
│   │   └── Tailgate or obtain PIN through observation
│   └── 2.4 Remove samples
│       └── Conceal in bag, exit normally
│
├── 3. Insider - Legitimate Access
│   ├── 3.1 Employee with freezer room access
│   ├── 3.2 Remove samples during normal work
│   └── 3.3 Conceal and exfiltrate
│       └── Personal bag, delivery packaging, etc.
│
├── 4. Insider - Exceeded Access
│   ├── 4.1 Employee without freezer access
│   ├── 4.2 Tailgate or social engineer into freezer room
│   └── 4.3 Remove samples
│
└── 5. Coerced Insider
    ├── 5.1 External actor identifies employee with access
    ├── 5.2 Coerces through threat, bribery, or ideology
    └── 5.3 Insider removes samples as directed

Path Analysis:

Path 1 (Forced Entry) is high effort and high detection risk. Modern alarm systems, camera coverage, and security officer response make this difficult. An attacker would need significant resources and would likely be detected before reaching the target.

Path 2 (Deception) is medium effort. The penetration test demonstrated this is feasible with moderate preparation. Key weaknesses are contractor verification and anti-tailgating.

Path 3 (Legitimate Access Insider) is the most dangerous because it’s nearly invisible. The insider uses existing access for unauthorized purposes. Detection depends on behavioral monitoring, inventory controls, and perhaps random inspections.

Path 4 (Exceeded Access Insider) combines insider knowledge with unauthorized access expansion. The insider knows what’s valuable but lacks access, so they use social engineering or tailgating to reach it.

Path 5 (Coerced Insider) is frightening because it can bypass most controls. The insider is legitimate, their behavior may appear normal, and their motivation is hidden.

Chokepoint Analysis:

The freezer room access point (Badge + PIN) appears in most attack paths. Strengthening this control disrupts multiple paths. Adding two-person rules, biometric verification, or behavioral monitoring at this point would significantly increase attack difficulty.

Sample inventory and chain of custody appear at the exfiltration stage. Even if an attacker reaches samples, robust inventory controls might detect the theft (though not prevent it).

Attack Tree 2: Sabotage BSL-3 Facility

Goal: Cause a pathogen release from the BSL-3 containment facility

BSL-3 Pathogen Release
├── 1. External Attack
│   ├── 1.1 Physical assault on building
│   │   ├── Vehicle ramming
│   │   ├── Explosive attack
│   │   └── Armed assault
│   └── 1.2 Remote sabotage
│       ├── Cyber attack on HVAC/containment systems
│       └── Utility disruption affecting containment
│
├── 2. Unauthorized Entry
│   ├── 2.1 Defeat all access controls (see previous analysis)
│   │   └── Six control points, three authentication factors
│   ├── 2.2 Navigate facility without proper PPE
│   │   └── Attacker risks self-exposure
│   └── 2.3 Breach containment
│       └── Open biosafety cabinets, damage HEPA systems
│
├── 3. Authorized Insider (Malicious)
│   ├── 3.1 Employee with BSL-3 access
│   ├── 3.2 Bypass or defeat containment procedures
│   │   └── Must defeat buddy system, cameras, behavioral training
│   └── 3.3 Release pathogens
│       └── Open containers outside BSC, damage seals, etc.
│
├── 4. Authorized Insider (Coerced)
│   ├── 4.1 External actor identifies BSL-3 personnel
│   ├── 4.2 Coerces through threat or manipulation
│   └── 4.3 Insider performs sabotage
│
└── 5. Negligent Insider
    ├── 5.1 Procedure violations compound
    ├── 5.2 Safety systems not functioning properly
    └── 5.3 Accidental release during normal work

Path Analysis:

Path 1 (Physical assault or cyber attack) is dramatic but difficult. The BSL-3 facility is purpose-built for containment, with redundant systems designed to maintain negative pressure even during power failures. A vehicle ramming might damage the building exterior but wouldn’t immediately breach containment. Cyber attacks are possible but would require extensive reconnaissance and sophistication.

Path 2 (Unauthorized physical entry) is extremely difficult. The attacker must defeat six access control points, then navigate a specialized facility without proper training or equipment. They’d likely expose themselves before accomplishing anything.

Path 3 (Malicious insider) is the nightmare scenario. A trained employee with legitimate access knows exactly how to breach containment. The buddy system is the primary control, but motivated insiders can potentially arrange circumstances where they’re briefly alone. Behavioral monitoring and a strong safety culture are critical.

Path 4 (Coerced insider) adds external adversary capability to insider access. This is particularly concerning for nation-state threats.

Path 5 (Negligent insider) is actually the most likely release scenario statistically, but it’s a safety concern rather than a security threat. The distinction matters for response planning.

Kill Chain for Physical Intrusion

Physical attacks follow a predictable sequence. Understanding this sequence reveals intervention points.

Each phase is a potential interdiction point. Early detection provides more response options.

Stage 7: Risk and Impact Analysis

Risk Priority Summary

Business Impact Quantification

THR-016/THR-006: Proprietary Sample Theft or Destruction

Loss of the HX-401 biological sample repository would be devastating:

The critical samples are literally irreplaceable in some cases—they represent unique cell lines, microbial strains, or engineered constructs that took years to develop.

THR-001: Tailgating

Tailgating is an enabling threat rather than a terminal threat—it creates access for other attacks. The impact is the full range of what an unauthorized person might do once inside.

THR-032: Controlled Substance Diversion

Mitigation Plan

Immediate Actions (Week 1-2)

30-Day Actions

90-Day Actions

Long-Term Actions (6-12 Months)

Risk Treatment Decisions

Risk Acceptance: Vehicle-Borne Threat

A vehicle-based attack is theoretically possible but extremely unlikely given Axiom’s profile. We accept this risk with the following conditions:

• Monitor threat intelligence for biotech-sector targeting
• Review if facility profile changes (controversial research, high-profile negative attention)
• Basic perimeter provides some standoff distance

Residual Risk Assessment

Summary: Critical threat count drops from 3 to 1 (insider sabotage remains challenging). High threat count drops from 10 to 4.

Budget Summary

For a company with $350M in facility investment and billions in potential research value, $3.2M in security investment (less than 1% of physical asset value) represents prudent protection.

Key Findings Summary

Physical Security Principles

This threat model illustrates principles specific to physical security:

Principle 1: Defense in Depth with Zones

Multiple security zones with increasing controls force attackers to defeat multiple systems. Each zone buys time for detection and response.

Principle 2: The Insider Advantage

Insiders bypass most physical controls by design—that’s what access is for. Insider threat mitigation requires behavioral controls, two-person rules, and detection-oriented thinking.

Principle 3: Access + Knowledge + Motivation = Threat

A threat requires all three elements. Remove access through controls. Limit knowledge through compartmentalization. Address motivation through personnel security and culture.

Principle 4: Humans Are Both Strongest and Weakest Controls

An alert security officer catches what cameras miss. A distracted employee holds the door for an attacker. Security culture training isn’t optional.

Principle 5: Response Time Matters

Physical attacks take time. The gap between detection and attacker success is the response window. Controls should maximize that window.

Principle 6: Prevention, Detection, and Response Are All Essential

You can’t prevent everything. You can’t detect everything instantly. You can’t respond to everything successfully. But layering all three creates a resilient security posture.

Comparison with Digital Threat Models

Physical and digital security are increasingly connected. Access control systems are networked. Laboratory equipment has software. A cyber attack could disable physical security controls, and physical access could enable cyber attacks. Comprehensive security requires addressing both domains.

Part 3e Complete | Physical Security Threat Model